Archived community.zenoss.org | full text search
Skip navigation
Currently Being Moderated

IRC September 2011

VERSION 1 
Created on: Dec 12, 2011 1:35 PM by Nick Yeates - Last Modified:  Dec 12, 2011 1:35 PM by Nick Yeates

<[01-Sep-2011 00:00:01] [connected at Thu Sep  1 00:00:01 2011]

[01-Sep-2011 00:00:01] [disconnected at Thu Sep  1 00:00:01 2011]

[01-Sep-2011 00:00:02] [connected at Thu Sep  1 00:00:02 2011]

[01-Sep-2011 00:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[01-Sep-2011 00:21:15] willwh`afk is now known as willwh

[01-Sep-2011 00:30:40] <tvincent__> snmp uses oid's

[01-Sep-2011 01:23:13] <tvincent__> anyone alive tonight?

[01-Sep-2011 01:24:06] <tvincent__> How do I add a event transformation? I want to add a transformation for net to show percentages. Any ideas?

[01-Sep-2011 01:53:57] willwh is now known as willwh`afk

[01-Sep-2011 01:54:07] willwh`afk is now known as willwh

[01-Sep-2011 02:09:57] willwh is now known as willwh`afk

[01-Sep-2011 04:17:07] <kotique> Now Zenoss's alerting in wrong schedule period.

[01-Sep-2011 04:17:43] <kotique> is this possible? we had proper schedules and proper alerting, but after some manipulations, it stopped doing that right.

[01-Sep-2011 04:49:20] hmp_ is now known as hmp

[01-Sep-2011 08:03:44] <konvert> hello

[01-Sep-2011 08:05:11] <konvert> who can help me with a problem with zenoos

[01-Sep-2011 08:05:16] <konvert> *zenoss

[01-Sep-2011 08:05:20] <Sashness> whats the problem

[01-Sep-2011 08:06:23] <konvert> i have problems with some traps zenoss doesn't translate

[01-Sep-2011 08:06:37] <konvert> i know i have to upload the specific mib files from the vendor.

[01-Sep-2011 08:07:05] <konvert> but i now have the mib files, i know how to upload the files with the command line in zenoss

[01-Sep-2011 08:07:34] <konvert> but after uploading the mib files, the traps aren't translated.

[01-Sep-2011 08:07:57] <konvert> I also watched in the mib files and didn't found the specific traps the machine sends.

[01-Sep-2011 08:08:19] <konvert> what can be the problem?

[01-Sep-2011 08:10:49] <jmp242> konvert: it's possible the MIB file is crap

[01-Sep-2011 08:11:04] <jmp242> or that you don't actually have the MIB file that defines the specific trap you're getting

[01-Sep-2011 08:11:22] <jmp242> some vendors like having 5 different MIB files for one device for some reason

[01-Sep-2011 08:11:57] <jmp242> What you can do is use an event transform in Zenoss to make the event say whatever you want (assuming you actually know what the trap is trying to say)

[01-Sep-2011 08:13:48] <konvert> so but i have for example this trap from a 3com device

[01-Sep-2011 08:13:49] <konvert>  snmp trap 1.3.6.1.4.1.43.45.1.2.23.1.14.1

[01-Sep-2011 08:14:07] <konvert> that whole list of numbers has to be in the mib files or not?

[01-Sep-2011 08:14:58] <jmp242> Mmmm, I don't actually know MIB files that well . . . but I think it has to be able to resolve that whole number yes

[01-Sep-2011 08:15:11] <jmp242> however, it can use multiple MIB files to do so through something like inheritance

[01-Sep-2011 08:17:06] <konvert> ok

[01-Sep-2011 08:17:36] <jmp242> usually if it can somewhat resolve the MIB it will do so though.

[01-Sep-2011 08:17:53] <jmp242> That's definately a 3com mib, but  1.3.6.1.4.1.43 = 3com

[01-Sep-2011 08:18:11] <jmp242> but the .45 I can't find any reference for that tree existing from the various quick google searches on MIBS on the net

[01-Sep-2011 08:18:21] <jmp242> I'm guessing the .45 isn't defined anywhere

[01-Sep-2011 08:19:25] <jmp242> wait, netdisco site says it might be a port # for a 48 port switch?

[01-Sep-2011 08:20:14] <konvert> no only 16 + 4 ports

[01-Sep-2011 08:25:32] <jmp242> hmm

[01-Sep-2011 08:25:37] <jmp242> the MIB examples on the net

[01-Sep-2011 08:25:42] <jmp242> http://www.snmplink.org/cgi-bin/nd/m/Ent/0-9/3Com/%5BALL%5D/jvco/a3com-huawei-device.mib

[01-Sep-2011 08:25:50] <jmp242> does that look like your MIB?

[01-Sep-2011 08:25:52] <jmp242> cause it doesn'

[01-Sep-2011 08:25:53] <tsener> anyone knows how can I monitor disk io utilization on a windows machine ?

[01-Sep-2011 08:26:29] <jmp242> provide any "translation" for the OID that I can see

[01-Sep-2011 08:27:32] <jmp242> tsener: do you use any of the Windows Zenpacks

[01-Sep-2011 08:27:54] <jmp242> cause the Egor one, under disk, provides IO Bytes and Disk Queue for me

[01-Sep-2011 08:27:58] <tsener> I use the Dell HW specific zenpack

[01-Sep-2011 08:28:12] <tsener> do you have a link for this one ?

[01-Sep-2011 08:29:14] <jmp242> docs/DOC-3500

[01-Sep-2011 08:29:21] <jmp242> watch the prereqs

[01-Sep-2011 08:47:23] <Hackman238> Morning all!

[01-Sep-2011 08:53:09] <jmp242> Morning

[01-Sep-2011 09:01:16] <konvert> jmp242 in your example it is correct until the last part

[01-Sep-2011 09:01:38] <konvert> .1.3.6.1.4.1.43.45.1.2.23.1 thats correct but then my trap is 14 in stead of 18

[01-Sep-2011 09:06:53] <jmp242> If you open your MIB file, can you find that OID?

[01-Sep-2011 09:10:28] * Sam-I-Am begins coffee

[01-Sep-2011 09:11:22] * Simon4 ends run and begins protein

[01-Sep-2011 09:18:33] <Sam-I-Am> pffft

[01-Sep-2011 09:18:37] <Sam-I-Am> healthy :/

[01-Sep-2011 09:19:39] <Simon4>

[01-Sep-2011 09:23:58] <Hackman238> Sam-I-Am: Coffee FTW!

[01-Sep-2011 09:25:31] <konvert> no can't find it

[01-Sep-2011 09:25:43] <jmp242> There's your problem ke4qqq

[01-Sep-2011 09:25:44] <jmp242> err

[01-Sep-2011 09:25:46] <jmp242> konvert:

[01-Sep-2011 09:25:49] <jmp242> damn autocomplete

[01-Sep-2011 09:26:00] <jmp242> if your MIB doesn't have it, Zenoss can't translate it

[01-Sep-2011 09:26:06] <jmp242> I suggest using a transform

[01-Sep-2011 09:27:19] <konvert> what do u mean with a transform

[01-Sep-2011 09:29:33] <jmp242> an event transform in zenoss

[01-Sep-2011 09:29:38] <jmp242> assuming you know what the trap means

[01-Sep-2011 09:36:16] <konvert> ok will try to figure out what the traps mean and then make an event transform

[01-Sep-2011 09:37:09] <konvert> something else, what does this mean ?

[01-Sep-2011 09:37:10] <konvert> snmp trap snmp_authenticationFailure

[01-Sep-2011 09:37:16] <konvert> i get it on some of my machines

[01-Sep-2011 09:39:17] <Hackman238> konvert: depending on how complex the mibs are you might be able to clean up the mibs by hand

[01-Sep-2011 09:39:22] <Sam-I-Am> someone or something had a login problem

[01-Sep-2011 09:39:30] <Sam-I-Am> perhaps another snmp client

[01-Sep-2011 09:39:54] <Hackman238> konvert: Or write a python script to read them and create a objexts.xml in zope xml markup to import the mibs directly as objects

[01-Sep-2011 09:42:27] <konvert> Hackman: that would be way to complex for me

[01-Sep-2011 09:42:46] <konvert> Sam-I-Am: zenoss is our only snmp receiver

[01-Sep-2011 09:43:08] <Sam-I-Am> incorrect community string somewhere?

[01-Sep-2011 09:44:20] <konvert> Sam-I-Am: what does this trap mean? that the community string i use in zenoss is not correct, or that some other device tries to read the snmp with the wrong community string?

[01-Sep-2011 09:45:13] <Sam-I-Am> either or... you're just getting the trap for it

[01-Sep-2011 09:45:30] <Sam-I-Am> the device was accessed with an incorrect community string

[01-Sep-2011 09:45:43] <Sam-I-Am> or the community string didnt have access to certain oids

[01-Sep-2011 09:47:34] <konvert> Sam-I-Am: and how to resolve the last problem, because i use the right community string an no other devices are trying to connect with snmp

[01-Sep-2011 09:48:00] <Hackman238> konvert: What are the mibs for?

[01-Sep-2011 09:56:36] kocolosk is now known as kocolosk|away

[01-Sep-2011 09:57:28] <konvert> Hackman238: for translating some traps for some of my machines

[01-Sep-2011 09:58:58] <Hackman238> konvert: Are there many? What type of device? I may have pretranslated ones I can share.

[01-Sep-2011 10:00:08] <konvert> Hackman238: ok here they come

[01-Sep-2011 10:00:33] <konvert> Kyocera FS-1118mfp (printer)

[01-Sep-2011 10:00:50] <Hackman238> konvert: Hum- thats a tough one

[01-Sep-2011 10:01:08] <Hackman238> konvert: Would you mind emailing me the mibs to take a peek at? hackman238@gmail.com

[01-Sep-2011 10:01:18] <Hackman238> konvert: or shane.scott@rackspace.com

[01-Sep-2011 10:01:33] <SDuensin> Greetings.

[01-Sep-2011 10:01:45] <Hackman238> SDuensin: Hey

[01-Sep-2011 10:03:47] <SDuensin> Hey Hackman238

[01-Sep-2011 10:03:52] <SDuensin> Oh.  I still hate Windows.

[01-Sep-2011 10:04:07] <Hackman238> SDuensin: WMI still giving you gripe?

[01-Sep-2011 10:04:21] <konvert> Hackman238: typing the mail atm

[01-Sep-2011 10:04:35] <Hackman238> SDuensin: I've been tinkering with WMI against Win2K8

[01-Sep-2011 10:05:02] <SDuensin> More than WMI.  I made an entire "package" we're pushing out.  Got it working under XP in a domain.  *Almost* working on W7.

[01-Sep-2011 10:05:19] <Hackman238> SDuensin: Ah thats right- the remote execution problem

[01-Sep-2011 10:05:46] <Hackman238> SDuensin: Still a token problem?

[01-Sep-2011 10:06:25] <SDuensin> Past that.  Now I've got the Cygwin SSHD not installing because it tells me /var/run needs to be writable by both whoever installed SSHD and Administrators.  But it is.  :-/

[01-Sep-2011 10:06:43] <konvert> Hackman238: mail send

[01-Sep-2011 10:07:29] <Hackman238> konvert: Gotcha- and you need to figure what these traps mean and put in transforms?

[01-Sep-2011 10:07:38] <Hackman238> SDuensin: Oh fun.

[01-Sep-2011 10:09:55] <SDuensin> Yea.  So I'm *hoping* it's a POSIX ACL that's wrong.  Otherwise, Cygwin is f'ed in the head and I'll have to figure out why.

[01-Sep-2011 10:10:16] <konvert> Hackman238: how can i figure this out ?

[01-Sep-2011 10:13:04] <Hackman238> SDuensin: Good luck.

[01-Sep-2011 10:13:10] <SDuensin> Hackman238, no kidding.

[01-Sep-2011 10:13:13] <Hackman238> SDuensin: Does seem like a ACL

[01-Sep-2011 10:13:23] <Hackman238> SDuensin: Pray that it isnt cygwin

[01-Sep-2011 10:13:25] <Hackman238> LOL

[01-Sep-2011 10:13:39] <SDuensin> Coffee first.  Cygwin second.

[01-Sep-2011 10:13:45] <Hackman238> konvert: Do you know what each of these traps mean already?

[01-Sep-2011 10:14:01] <Hackman238> SDuensin: Always- caffination is a prereq to ahrd work

[01-Sep-2011 10:15:53] <konvert> no i want to know how i can figure this out?

[01-Sep-2011 10:16:15] <Hackman238> konvert: Ah gotcha. I'd hit google first. I'll glance in just a minute

[01-Sep-2011 10:16:27] <Hackman238> Short on threads.

[01-Sep-2011 10:19:42] <kickehy> So, I seem to be getting a couple heartbeat failures on my zenoss box, what's the best way to troubleshoot those?

[01-Sep-2011 10:21:59] <froztbyte> wut. I've got something /Device/Linux that's not inheriting the ethernetCsmacd perf template :/

[01-Sep-2011 10:26:39] <Hackman238> konvert: <ModelOID>1.3.6.1.4.1.1347.43.5.1.1.1.1</ModelOID>

[01-Sep-2011 10:27:15] <Hackman238> konvert: 1.3.6.1.4.1.1347.43.5.1.1.1.2 -> ?

[01-Sep-2011 10:28:49] <Hackman238> konvert: WCIT_OID_hwdot1sStpStatus      => ".1.3.6.1.4.1.43.45.1.2.23.1.14.1.0"

[01-Sep-2011 10:29:46] <Hackman238> konvert: 1.3.6.1.4.1.6101.1.8.8.2.4    Violation of the URL-blocking and URL Filtering rules

[01-Sep-2011 10:30:08] <Hackman238> konvert: 1.3.6.1.4.1.6101.1.8.8.2.5    Scan components (pattern, engine) update notification

[01-Sep-2011 10:32:34] <Hackman238> froztbyte: Are there any local overrides?

[01-Sep-2011 10:32:48] <Hackman238> kickehy: Which daemons?

[01-Sep-2011 10:38:34] <froztbyte> Hackman238: no, not in any of the paths

[01-Sep-2011 10:38:43] <froztbyte> Hackman238: which is why it's downright strange

[01-Sep-2011 10:39:45] <froztbyte> it's under /Devices/Server/Linux/Prx. I see it all the way to .../Linux (evident in those devices getting polled and perf-data tracked)

[01-Sep-2011 10:39:54] <froztbyte> but .../Linux/Prx has nada

[01-Sep-2011 10:40:17] <froztbyte> I think I should just ignore it and look at it again tomorrow

[01-Sep-2011 10:42:45] <Hackman238> froztbyte: Very strange,

[01-Sep-2011 10:42:58] <kickehy> Hackman238: sorry for the delay, zenstatus and zenmodeler

[01-Sep-2011 10:43:08] <froztbyte> Hackman238: yeah, I'm totally just gonna ignore it until tomorrow

[01-Sep-2011 10:43:21] <froztbyte> Hackman238: it's 16h43 here anyway, so it's a safe strategy

[01-Sep-2011 10:43:25] <Hackman238> froztbyte: Try a zenchkrels -x1 -r before you leave work

[01-Sep-2011 10:43:40] <Hackman238> froztbyte: Just incase theres a rrdTemplate rel issue

[01-Sep-2011 10:43:46] * froztbyte never really leaves work

[01-Sep-2011 10:43:51] <Hackman238> froztbyte: Sorry

[01-Sep-2011 10:43:51] <froztbyte> Hackman238: but yeah, that's not a bad idea

[01-Sep-2011 10:43:57] <Hackman238> kickehy: NP

[01-Sep-2011 10:45:35] <jmp242> Hackman238: this looks like some sort of timeout setting to me . . . what do you think? message/60960#60960

[01-Sep-2011 10:47:18] <froztbyte> Hackman238: I'm actually partial to thinking something's wrong with .../Prx

[01-Sep-2011 10:47:28] <froztbyte> since it doesn't look like /anything/ there gets run

[01-Sep-2011 10:48:04] <froztbyte> luckily it's only two servers under that node on this instance, so I can nuke & redo it

[01-Sep-2011 10:50:00] <tsener> It seems there is a freebsd port released

[01-Sep-2011 10:50:07] <tsener> anyone tried it ?

[01-Sep-2011 10:50:13] <tsener> its 3.1 version

[01-Sep-2011 10:53:17] <Hackman238> jmp242: Got it. TY

[01-Sep-2011 10:54:15] <Hackman238> froztbyte: try method two here: http://shanewilliamscott.com/Code/zendmd-tip-fix-broken-relationships.html

[01-Sep-2011 10:54:44] <Hackman238> froztbyte: Keep note of anything returned during the op. A clean run wont return anything

[01-Sep-2011 10:55:40] <Hackman238> froztbyte: Also you can substute d.getsubObjects() with dmd.Devices.Servers.YourClass.getSubObjects() to check specific areas

[01-Sep-2011 10:55:40] <froztbyte> Hackman238: screen + zendmd + time

[01-Sep-2011 10:55:42] <froztbyte>

[01-Sep-2011 10:55:47] <Hackman238> froztbyte: Aye

[01-Sep-2011 10:55:49] <Hackman238>

[01-Sep-2011 10:55:58] <froztbyte> start it now, wait until later

[01-Sep-2011 10:56:03] <Hackman238> tsener: I havent, sorry

[01-Sep-2011 10:56:05] <froztbyte> and I'm skipping the commit for now

[01-Sep-2011 10:56:33] <Hackman238> froztbyte: Sure, np

[01-Sep-2011 10:56:49] <froztbyte> actually

[01-Sep-2011 10:56:56] * froztbyte makes a backup, just for good measure

[01-Sep-2011 10:57:34] <Hackman238> froztbyte: Never hurts.

[01-Sep-2011 10:57:45] <Hackman238> froztbyte: Do a zeopack- best backup

[01-Sep-2011 11:00:49] <mrchippy> hola, everybody

[01-Sep-2011 11:02:01] <user3322> Pryvet

[01-Sep-2011 11:03:21] <Hackman238> mrchippy: Hey, how goes it?

[01-Sep-2011 11:03:47] <mrchippy> it goes crazy awesome

[01-Sep-2011 11:04:19] <Hackman238> user3322: ?

[01-Sep-2011 11:04:24] <Hackman238> mrchippy: Glad to ehar

[01-Sep-2011 11:05:06] <mrchippy> privyet is russian for howdy

[01-Sep-2011 11:05:08] <user3322> i thought it was "say hello in a different language day" so I used Russian…

[01-Sep-2011 11:05:28] <Hackman238> mrchippy: Ah gotcha

[01-Sep-2011 11:05:48] <Hackman238> user3322: guten tag

[01-Sep-2011 11:05:49] <nyeates> Hi all

[01-Sep-2011 11:06:07] <Hackman238> nyeates: Hey, how goes the war?

[01-Sep-2011 11:07:27] <nyeates> Ok Hackman238... got any stumpers for mrchippy today?

[01-Sep-2011 11:07:30] <rmatte> tsener: I doubt anyone in here has tried the BSD port... bad experiences with running Zenoss on BSD in the past

[01-Sep-2011 11:08:22] <nyeates> The guys at Experts exchange, i think, have made new BSD-compliant code

[01-Sep-2011 11:09:00] <Hackman238> mrchippy: Any idea how I can get my heartbeats to work more reliably in v2.5.2? Most of my DC's have hb issues all the time even though all the daemons are working away flawlessly

[01-Sep-2011 11:09:07] <nyeates> ahh freeBSD: http://www.experts-exchange.com/blogs/EE-Tech-News/B_5409-Experts-Exchange-Employee-Makes-It-Possible-to-Install-Zenoss-on-a-FreeBSD-Machine.html

[01-Sep-2011 11:09:43] <Hackman238> mrchippy: Also, where can I find technical documentation on how the failover for collectors works in Zenoss 4.0.2?

[01-Sep-2011 11:11:08] <nyeates> see here for freeBSD port info: message/60551

[01-Sep-2011 11:12:00] <mrchippy> on the reliable heartbeats, i can't tell you a good answer other than improvements were made in that area in 3 and especially 4.   Sorry if that's an asshat answer If you have a more specific situation, I might be able to dig out an answer for you

[01-Sep-2011 11:13:10] <nyeates> Hackman238: one of yer production instances still on 2.5.2? understandable - just asking

[01-Sep-2011 11:13:34] <mrchippy> on the tech docs for collector failover in 4.0.2-- what is this "collector failover" you speak of?

[01-Sep-2011 11:13:43] <Hackman238> nyeates: All of our production is on 2.5.2....we're waiting on Zenoss professional services to fix a few packs in the contracts...stil

[01-Sep-2011 11:13:56] <Hackman238> mrchippy: 1 sec, I'll find the listing

[01-Sep-2011 11:14:03] <mrchippy> cool, thx

[01-Sep-2011 11:14:44] <Hackman238> mrchippy: http://www.google.com/search?q=zenoss+collector+failover&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a#sclient=psy&hl=en&client=firefox-a&hs=hHp&rls=org.mozilla:en-US%3Aofficial&biw=1680&bih=834&source=hp&q=zenoss+Collector+redundancy+&pbx=1&oq=zenoss+Collector+redundancy+&aq=f&aqi=&aql=1&gs_sm=e&gs_upl=21199l21225l0l21976l3l2l0l0l0l0l223l290l1.0.1l2l0&bav=on.2,or.r_gc.r_pw.&fp=af45d49d6de0dc92

[01-Sep-2011 11:14:45] <konvert> Hackman238 thx

[01-Sep-2011 11:14:55] <Hackman238> mrchippy: "Collector redundancy'

[01-Sep-2011 11:15:25] <rmatte> Hackman238: join the club, constant heartbeat issues in 2.5.2

[01-Sep-2011 11:15:31] <Hackman238> konvert: NP

[01-Sep-2011 11:15:33] <rmatte> Hackman238: I just ignore them at this point

[01-Sep-2011 11:16:21] <Hackman238> rmatte: yeah its a pain in the balls. We have huge caches, dedicated hubs, insanely fast mysql setups, cached configs true....can't figure it out

[01-Sep-2011 11:16:52] <Hackman238> rmatte: Whats really a pest is that all the DC's are configured exactly the same....but some have the issue, some dont.

[01-Sep-2011 11:17:12] <Hackman238> rmatte: ...and one in a while they'll go all musical chairs and some stop andothers start up with the problem.

[01-Sep-2011 11:17:20] <Hackman238> s/one/once/

[01-Sep-2011 11:17:41] <Hackman238> The madness! Madness!

[01-Sep-2011 11:17:56] <jmp242> lol

[01-Sep-2011 11:18:17] * Sam-I-Am had heartbeat issues too

[01-Sep-2011 11:18:42] <mrchippy> ahhh, ok that's our HA config that ps uses.  We have a doc on it.  Let me find it.

[01-Sep-2011 11:19:13] <Hackman238> mrchippy: Thanks. I've got folks after me for an explaination and all I can say at this point is 'no one at Zenoss has ever heard of the feature'

[01-Sep-2011 11:19:29] <nyeates> mrchippy: HA as in like DRBD?

[01-Sep-2011 11:19:30] <rmatte> Hackman238: heartbeat events == the boy who cried wolf

[01-Sep-2011 11:19:55] <Hackman238> rmatte: Aye...and you never know which DC is the hypocondriact and which arent!

[01-Sep-2011 11:20:01] <rmatte> lol

[01-Sep-2011 11:20:07] <mrchippy> yeah, it's not a 'feature' as in something built into the product.  it's a set of configuration instructions.

[01-Sep-2011 11:20:35] <Hackman238> s/hypocondriact/hypochondriac/

[01-Sep-2011 11:21:11] <bigegor> hi all

[01-Sep-2011 11:21:20] <Hackman238> mrchippy: Careful with that type of advertising. That means that monitoring the air fuel ratio using zenoss command datasources on a car pc is also a feature

[01-Sep-2011 11:21:21] <nyeates> mrchippy: i can provide link to the KB for "How To Configure Zenoss for High-Availability Using DRBD and Heartbeat"

[01-Sep-2011 11:21:38] <nyeates> Heartbeat in this sense being the linux kind

[01-Sep-2011 11:21:54] <Hackman238> nyeates: TY

[01-Sep-2011 11:22:34] <nyeates> Gimme a few minutes tho...its on the new KB and Its hard to provide links to it

[01-Sep-2011 11:22:49] <Hackman238> nyeates: NP man, no rush

[01-Sep-2011 11:22:56] <Hackman238> nyeates: Even later is good

[01-Sep-2011 11:23:04] <Hackman238> mrchippy: Whats your take on message/60974#60974

[01-Sep-2011 11:23:53] <konvert> Guys i have another question, we are running citrix, is there a zenpack for citrix?

[01-Sep-2011 11:24:37] <bigegor> i'm playing now with Zenoss 3.2 beta. It seems like something was changed in ZenCollector framework (new task creation). My question is Products.ZenCollector.tasks.SubConfigurationTaskSplitter class usable for now or it has WIP status.

[01-Sep-2011 11:27:03] <mrchippy> sorry i'm taking so long.  trying to digest the forum entry...

[01-Sep-2011 11:27:29] <Hackman238> mrchippy: NP. No rush.

[01-Sep-2011 11:29:06] <Hackman238> mrchippy: Also, bit of a dev question, is it possible to pass operands like 'AND' and 'OR' to the Zope catalog search?

[01-Sep-2011 11:29:37] <mrchippy> yeah.  Check out AdvancedQuery.  i'll find the link and post

[01-Sep-2011 11:29:45] <Hackman238> mrchippy: TY

[01-Sep-2011 11:30:00] <mrchippy> http://www.dieter.handshake.de/pyprojects/zope/AdvancedQuery.html

[01-Sep-2011 11:30:31] <Hackman238> mrchippy: Danke.

[01-Sep-2011 11:30:35] <mrchippy> np

[01-Sep-2011 11:31:00] <Hackman238> mrchippy: Who are you, by the way? Simon?

[01-Sep-2011 11:32:03] <mrchippy> my name is Chip Holden.  i provide soda and snacks for the wieners in the engineering group

[01-Sep-2011 11:32:21] <mrchippy> i code some too

[01-Sep-2011 11:33:09] <nyeates> Public link to the HA / DRBD stuff btw... https://support.zenoss.com/ics/support/default.asp?deptID=16059&task=knowledge&questionID=8

[01-Sep-2011 11:33:56] <Hackman238> mrchippy: LOL. Gotcha. I like to know who's who so I can inform Rusty of who needs a raise for excellent work.

[01-Sep-2011 11:34:09] <Hackman238> nyeates: Thanks Nick

[01-Sep-2011 11:34:51] <dhopp> if I'm creating a distributed collector in core, rather then trying to install all zenpacks, can I just rsync $ZENHOME/zenpacks from the hub to the distributed collector and then restart zenoss on the collector?  (I know I'd have to grab /bin/* and /etc/*.conf as well and modify confs)

[01-Sep-2011 11:35:24] <rmatte> dhopp: yes, you can, though I'm pretty sure that Egor's ZenPack takes care of doing the ZenPack syncing

[01-Sep-2011 11:35:58] <dhopp> rmatte: I know his zenpack does handle it..but his zenpack doesn't handle logical collectors right now (running multiple collectors on the same physical host)

[01-Sep-2011 11:36:08] <rmatte> ah, yeh it doesn't

[01-Sep-2011 11:36:12] <dhopp> so I was just making sure it was that simple

[01-Sep-2011 11:36:16] <rmatte> but yeh, you're correct that you just have to rsync or copy

[01-Sep-2011 11:36:24] <mrchippy> dhopp: trying to think if there's location-specific stuff in the deployed zenpacks.... nothing coming to mind.

[01-Sep-2011 11:36:26] <nyeates> bigegor: is that issue halting your zenpack from working in 3.2?

[01-Sep-2011 11:37:02] <rmatte> dhopp: the only thing that matters is that any scripts that the collector may be trying to execute are there, meaning the ZenPacks don't actually have to be installed (since Zope isn't even running)

[01-Sep-2011 11:37:24] <dhopp> mrchippy:  I know some zenpacks that create a new daemon have the daemon live in the zenpack dir and then create a symlink in bin..but I can script around that to handle that scenario

[01-Sep-2011 11:37:31] <dhopp> rmatte:  that's what I thought..but wanted to verify

[01-Sep-2011 11:37:48] <rmatte> you're right about daemons though, that's a special case

[01-Sep-2011 11:37:59] <rmatte> and like you said, just need to create the symlink

[01-Sep-2011 11:38:08] <bigegor> i have some strange results (error: "Task servername already exist")

[01-Sep-2011 11:38:32] <Hackman238> bigegor: Strange indeed. Where do you see that?

[01-Sep-2011 11:40:09] <nyeates> bigegor: this is for your ZenPacks.community.DistributedCollectors-1.7.egg zenpack?

[01-Sep-2011 11:40:38] <bigegor> zenperfsql ZenPerfSqlTaskSplitter splitConfiguration method

[01-Sep-2011 11:41:19] <bigegor> nyeates: no, my new SQLDataSource

[01-Sep-2011 11:41:47] <jmp242> Yum, new forum thread that I've perhaps gone down the garden path on, or can't help anymore

[01-Sep-2011 11:41:48] <jmp242> message/60979#60979

[01-Sep-2011 11:42:14] <mrchippy> jmp242: looking

[01-Sep-2011 11:43:29] <jplouis> bigegor: do you have a link to your code?

[01-Sep-2011 11:44:17] <bigegor> jplouis: https://github.com/epuzanov/ZenPacks.community.SQLDataSource

[01-Sep-2011 11:47:04] <bigegor> jplouis: today i've played with Products.ZenCollector.tasks.SubConfigurationTaskSplitter class. It seems like it subconfiguration not fully supported by scheduler (error: "Task servername already exist").

[01-Sep-2011 11:47:12] <rocket> Hackman238: regarding ha there has been an KB article for it for ages on our old portal

[01-Sep-2011 11:48:04] <nyeates> rocket: the link to it on the new KB is above

[01-Sep-2011 11:52:22] <nyeates> welcome back davetoo

[01-Sep-2011 11:52:36] <nexex> Monitoring using JMX, if the java processes on the target is restarted, zenoss will say: DataSource Thread Count; Error calling mbean java.lang:type=Threading: Exception: no MBean registered with name: java.lang:type=Threading

[01-Sep-2011 11:52:49] <nexex> it will say that until I connect with jconsole, then it self clears

[01-Sep-2011 11:52:58] <davetoo> hi

[01-Sep-2011 11:53:01] <davetoo> busy, busy, busy

[01-Sep-2011 11:54:10] <dhopp> another question regarding redundancy.  I have a hub server in 1 DC and then thinking of a standby in a 2nd DC.  DRBD isn't really feasible in this situation.  The remote collectors will have hubhost set to a CNAME and in the event of a complete catastrophe I could change that CNAME to the hubhost in the 2nd DC.  I'm going to be rsyncing $ZENHOME from primary hub to secondary hub and will have mysql replication setup.  Is this something that *should* work or am

[01-Sep-2011 11:54:25] <jplouis> bigegor: We use SubConfigurationTaskSplitter in a few daemons. Probably some sort of undocumented setup needs to happen

[01-Sep-2011 11:55:46] <hmp> dhopp: why not use DRBD?

[01-Sep-2011 11:56:43] <rocket> dhopp: you have to worry about the Data.fs mostly

[01-Sep-2011 11:56:56] <rmatte> hmmm, so I'm trying to load some Mibs and I get an error: IOError: [Errno 2] No such file or directory: '/tmp/mib_extract/MrSEel/mib-jnx-l2ald.txt'

[01-Sep-2011 11:57:07] <rmatte> but, that file is in the directory that I'm trying to load the Mibs from

[01-Sep-2011 11:57:14] <rocket> dhopp: you will need to use repozo to create a backup regularly and then sync that across

[01-Sep-2011 11:57:26] <rmatte> so I don't get why it doesn't get copied to the tmp directory for installation if that's what's required

[01-Sep-2011 11:57:28] <rocket> rmatte: dir perms?

[01-Sep-2011 11:57:36] <rmatte> doubt it but I'll check

[01-Sep-2011 11:57:39] <dhopp> rocket: repozo?

[01-Sep-2011 11:57:53] <rocket> check the source code for zenbackup

[01-Sep-2011 11:58:08] <rocket> you will see that is the utility being used to backup the data.fs file to a new file

[01-Sep-2011 11:58:10] <dhopp> hmp: too many firewalls and other security that would have to be messed with to make that work

[01-Sep-2011 11:58:15] <rmatte> nope, I was able to copy it find from hand

[01-Sep-2011 11:58:29] <rmatte> fine* by hand

[01-Sep-2011 11:58:41] <rocket> since the Data.fs file is live you cannot just copy it between systems or it will be corrupt

[01-Sep-2011 11:58:44] <Simon4> dhopp: we do pretty much exactly that, and I have a cron script that syncs the Data.fs file in the way rocket describes every 15 minutes

[01-Sep-2011 11:59:02] <Simon4> so you have a cold standby install in the other DC

[01-Sep-2011 11:59:11] <nyeates> repozo is  zope tool for incremental backup w/o having to shut off zopectl

[01-Sep-2011 11:59:19] <Hackman238> dhopp: Best bet to keep a backup of Data.fs is to zeopack it then rsyn the Data.fs.old to Data.fs on the other side

[01-Sep-2011 11:59:42] <Hackman238> dhopp: Since it is alfterall standby

[01-Sep-2011 11:59:47] <Hackman238> (after all

[01-Sep-2011 12:00:01] [disconnected at Thu Sep  1 12:00:01 2011]

[01-Sep-2011 12:00:01] [connected at Thu Sep  1 12:00:01 2011]

[01-Sep-2011 12:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[01-Sep-2011 12:00:21] <rocket> thats another way to do it too .. basically the same thing with the added advantage that its packed .. :/

[01-Sep-2011 12:01:02] <Hackman238> dhopp: If you zeopack often it wont take long

[01-Sep-2011 12:01:14] <Hackman238> dhopp: So say you pack every 4 hours, at worst you'll be 4 hours behind

[01-Sep-2011 12:01:25] <Hackman238> dhopp: Plus the pack wont take long after the initial pack

[01-Sep-2011 12:02:52] <rmatte> http://fpaste.org/Ei7S/raw/

[01-Sep-2011 12:03:03] <rmatte> That's what I get when trying to install the Mibs

[01-Sep-2011 12:03:15] <rmatte> not really sure why it's tarring them then extracting them from tar

[01-Sep-2011 12:03:24] <davetoo> mibs, mibs, mibs,

[01-Sep-2011 12:03:27] <Hackman238> rmatte: For fun!

[01-Sep-2011 12:03:28] <davetoo> it's a box all full of mibs,

[01-Sep-2011 12:03:29] <davetoo> ....

[01-Sep-2011 12:03:40] <davetoo> oh

[01-Sep-2011 12:03:42] <rmatte> zenmib can be such a pain in the ass

[01-Sep-2011 12:03:45] <Hackman238> rmatte: Burn those CPU cycles!

[01-Sep-2011 12:03:50] * davetoo reminds self to look at rmatte's cisco mib pack

[01-Sep-2011 12:03:50] <rmatte> lol

[01-Sep-2011 12:04:06] <rmatte> yeh, getting those cisco Mibs loaded was a pain

[01-Sep-2011 12:04:10] <dhopp> Hackman238 / rocket: thanks…I didn't realize zeopack created a .old

[01-Sep-2011 12:04:42] <rmatte> Also, apparently zenmib was rewritten by Kells to be better, but instead it actually loads less of the Cisco Mibs now

[01-Sep-2011 12:04:57] <Hackman238> dhopp: NP

[01-Sep-2011 12:05:06] <davetoo> rmatte do you have a favorite standalone mib browser tool?  I used to use TkIned, but... it doesn't work with later versions of tcl/tk

[01-Sep-2011 12:05:12] <Hackman238> rmatte: LOL Did he jsut comment out the areas that were broke?

[01-Sep-2011 12:05:15] <rmatte> though it does fix a couple of problems where I had to modify some of the Mibs by hand to get them to load in

[01-Sep-2011 12:05:21] <Hackman238> rmatte: I'm JK, kells is good.

[01-Sep-2011 12:05:24] <rmatte> nah, doubt it

[01-Sep-2011 12:05:27] <rmatte> lol

[01-Sep-2011 12:06:04] <Hackman238> LOL

[01-Sep-2011 12:06:28] <davetoo> try: something; except: give_up()

[01-Sep-2011 12:06:46] <dhopp> Hackman238:  so I could so something like this on the secondary: rsync primary:/opt/zenoss —exclude=Data.fs; cp /opt/zenoss/var/Data.fs.old /opt/zenoss/var/Data.fs

[01-Sep-2011 12:06:53] <rmatte> davetoo: I honestly just browse the Mibs online at the oidview site

[01-Sep-2011 12:06:57] <rmatte> works just as well

[01-Sep-2011 12:07:04] <dhopp> or maybe do that from the primary right after a zeopack

[01-Sep-2011 12:07:24] <rmatte> http://www.oidview.com/

[01-Sep-2011 12:07:38] <davetoo> heh.. I'm trying to get permission to buy a copy of oidview; the automagic mib-downloading has already saved me hours, running the trial version.

[01-Sep-2011 12:07:53] <Hackman238> dhopp: yep

[01-Sep-2011 12:08:00] <rmatte> well, I just use the online version of it

[01-Sep-2011 12:08:06] <davetoo> yeah

[01-Sep-2011 12:08:10] <rmatte> if you google for a MIB name the first result is almost always oidview online

[01-Sep-2011 12:08:14] <Hackman238> dhopp: But you'll want to change your cron for the zeopack to be far more frequent

[01-Sep-2011 12:08:50] <Hackman238> dhopp: That works too

[01-Sep-2011 12:10:19] <rmatte> eugh, juniper sucks balls with Mibs, one of these mib files in completely empty

[01-Sep-2011 12:10:23] <davetoo> I get paid to deal with a different set of systems every week or two; anything that saves me time like this ought to be a no-brainer   I could spend a bunch of time fussing around with snmpwalk and manual mib management, but this is so much easier.

[01-Sep-2011 12:10:42] <davetoo> __init__.mib ? :

[01-Sep-2011 12:10:45] <Hackman238> rmatte: s/eugh, juniper sucks balls with Mibs/eugh, juniper sucks balls/

[01-Sep-2011 12:10:58] <rmatte> haha, that's what the problem was, empty Mib file packaged in

[01-Sep-2011 12:11:09] <rmatte> zenmib was choking on it

[01-Sep-2011 12:11:45] <rmatte> mystery solved

[01-Sep-2011 12:12:33] <rmatte> these are the Mibs for JunOS

[01-Sep-2011 12:12:37] <rmatte> hadn't loaded them in yet

[01-Sep-2011 12:12:44] <dhopp> Hackman238:  thanks..obviously if I do that on the primary I change the cp command slightly :-)

[01-Sep-2011 12:13:04] <Hackman238> rmatte: I think we should write some how to's and FAQ on MIBs on the ZCA wiki

[01-Sep-2011 12:13:17] <rmatte> Hackman238: well, there's really not much to know

[01-Sep-2011 12:13:33] <rmatte> Extract mibs to directory then: zenmib run -v10 *

[01-Sep-2011 12:13:34] <rmatte> lol

[01-Sep-2011 12:13:45] <Hackman238> rmatte: Just tips and tricks on getting zenmib to work.

[01-Sep-2011 12:13:47] <rmatte> If it fails, troubleshoot as best you can

[01-Sep-2011 12:13:49] <rmatte> then give up

[01-Sep-2011 12:13:54] <Hackman238> rmatte: well tricking zenmib in to working.

[01-Sep-2011 12:14:01] <rmatte> also need to restart Zenoss after the Mibs are installed

[01-Sep-2011 12:14:30] <rmatte> yeh, the problem is the tricks aren't exactly general

[01-Sep-2011 12:14:42] <rmatte> it's more about coming up with ways on the fly

[01-Sep-2011 12:15:04] <rmatte> voila, juniper mibs installed

[01-Sep-2011 12:15:12] <nyeates> Im going to change the topic and officiate the dev meeting over....thx everyone

[01-Sep-2011 12:15:17] <Hackman238> rmatte: True, but  not all users can troubleshoot by themselves LOL

[01-Sep-2011 12:15:19] <Hackman238> rmatte: Nice

[01-Sep-2011 12:15:27] <davetoo> heh.. just realized that

[01-Sep-2011 12:15:28] <Hackman238> nyeates: Thanks Nick!

[01-Sep-2011 12:15:35] <davetoo> was supposed to be in progresse

[01-Sep-2011 12:16:11] <rmatte> erm...

[01-Sep-2011 12:16:18] <rmatte> true, but you can't really document crafty troubleshooting

[01-Sep-2011 12:16:26] <rmatte> you can document common issues, but that's about it

[01-Sep-2011 12:16:36] <Hackman238> rmatte: Perhaps a webmd for zenoss!

[01-Sep-2011 12:16:49] <davetoo> it's .. surprisingly nondeterministic

[01-Sep-2011 12:16:50] <rmatte>

[01-Sep-2011 12:17:03] <Hackman238> LOL

[01-Sep-2011 12:17:03] <rmatte> "The doctor is in"

[01-Sep-2011 12:17:04] <davetoo> if it's not to be super-document-heavy.

[01-Sep-2011 12:17:39] * davetoo used to work for a NASA subcontractor on the space shuttle program (never got to see a launch).  I had to write down every single step that I took that could be described.

[01-Sep-2011 12:18:17] <nyeates> "next, i removed my bottom from my chair, and got coffee"

[01-Sep-2011 12:18:26] <rmatte> Here's our zenmib troubleshooting documentation: http://xkcd.com/627/

[01-Sep-2011 12:18:53] <rmatte> would need some slight modification

[01-Sep-2011 12:18:56] <Hackman238> LOL

[01-Sep-2011 12:19:05] <davetoo> seriously: google is a great tool.

[01-Sep-2011 12:19:08] <Hackman238> rmatte: Very slight modification

[01-Sep-2011 12:19:15] <Hackman238>

[01-Sep-2011 12:19:16] <rmatte> lol

[01-Sep-2011 12:19:21] <davetoo> except when all it finds is me asking the same question a year before, with no answer.

[01-Sep-2011 12:19:41] <rmatte> s/click/enter command and return/g

[01-Sep-2011 12:19:47] <rmatte> done

[01-Sep-2011 12:19:47] <nyeates> that xkcd is sooo true

[01-Sep-2011 12:19:57] <rmatte> nyeates: yeh, it really is

[01-Sep-2011 12:19:59] <davetoo> they forgot -v10

[01-Sep-2011 12:20:11] <rmatte> that too

[01-Sep-2011 12:20:16] <Hackman238> davetoo: true, need v10

[01-Sep-2011 12:20:25] <Hackman238> davetoo: if not only to look busy

[01-Sep-2011 12:20:28] <Hackman238> LOL

[01-Sep-2011 12:26:06] <rmatte> the best comics on the net though are here: http://theoatmeal.com/

[01-Sep-2011 12:27:10] <rmatte> For instance: http://theoatmeal.com/comics/apple

[01-Sep-2011 12:29:00] <davetoo> http://art.penny-arcade.com/photos/215181106_tvEcr-XL-2.jpg

[01-Sep-2011 12:29:02] <davetoo> that is me.

[01-Sep-2011 12:29:06] <davetoo> every day.

[01-Sep-2011 12:29:31] <rmatte> haha

[01-Sep-2011 12:29:37] <davetoo> Mostly the coffe part,

[01-Sep-2011 12:29:58] <davetoo> but I also talk like a sysadmin-sailor :P)

[01-Sep-2011 12:30:07] <rmatte> http://theoatmeal.com/comics/customer_service <--- "Why I'd rather be punched in the balls than call customer service."

[01-Sep-2011 12:30:53] <davetoo> heh...

[01-Sep-2011 12:31:18] <rmatte> "Groan loudly as if in mid-orgasm every time the support operator starts to speak.  It will make them uncomfortable - this is how the battle is won."

[01-Sep-2011 12:32:24] <davetoo> Skippy McSkipperSkip

[01-Sep-2011 12:32:34] <davetoo> ok, back to serious work....

[01-Sep-2011 12:32:41] <rmatte> lol

[01-Sep-2011 12:34:38] <rmatte> "Hello sir!  Can I get your first name, phone number, address, date of birth, favorite planet in the solar system, and least favorite african mammal.  I won't actually log this information, mind you, so you'll have to repeat it to every other operator I forward you to."

[01-Sep-2011 12:37:53] <Hackman238> rmatte: Oh my. The apple one is hilarious!

[01-Sep-2011 12:38:07] <rmatte> This one is amazing too: http://theoatmeal.com/comics/online_gaming

[01-Sep-2011 12:39:59] <Hackman238> rmatte: "freaky weird sey illegal in 7 states" LOL

[01-Sep-2011 12:40:07] <rmatte> lol

[01-Sep-2011 12:40:08] <Hackman238> s/sey/sex/

[01-Sep-2011 12:40:11] <Hackman238> LOL

[01-Sep-2011 12:40:22] <rmatte> I like the guys running at them "but Mommmmmm...."

[01-Sep-2011 12:40:23] <rmatte> lol

[01-Sep-2011 12:40:30] <Hackman238> lol

[01-Sep-2011 12:40:30] <rmatte> It's so true

[01-Sep-2011 12:41:53] <rmatte> http://theoatmeal.com/comics/airplane_layout

[01-Sep-2011 12:42:11] <rmatte> That is the best layout for a plane ever, I especially like the screaming infant storage kennel

[01-Sep-2011 12:47:46] <Hackman238> LOL

[01-Sep-2011 13:01:55] <Sam-I-Am> hehe, theoatmeal is rock

[01-Sep-2011 13:04:42] <Hackman238> Sam-I-Am: Indeed

[01-Sep-2011 13:05:49] <Thinh> hey is remotecollector easy to install on zenoss core 3.0?

[01-Sep-2011 13:11:27] <rmatte> Thinh: you're talking about core?

[01-Sep-2011 13:11:42] <rmatte> and I hope you mean 3.1, because 3.0 is buggy as all hell

[01-Sep-2011 13:11:43] <Thinh> yea

[01-Sep-2011 13:11:48] <Thinh> yea 3.1 or whatever the latest

[01-Sep-2011 13:12:04] <rmatte> docs/DOC-5861

[01-Sep-2011 13:12:10] <Thinh> i know in the past, 2.5 you need to do some backend stuff right

[01-Sep-2011 13:12:12] <rmatte> You can use that ZenPack to manage remote collectors

[01-Sep-2011 13:12:16] <rmatte> it's easy enough to do

[01-Sep-2011 13:12:39] <rmatte> there's some pre-requisite configuration to do for SSH, if you read through the comments on the ZenPack page I listed the steps to take in a few posts that I made

[01-Sep-2011 13:12:42] <Thinh> what do you need to install on the collector box

[01-Sep-2011 13:12:45] <Thinh> just the first rpm?

[01-Sep-2011 13:12:47] <rmatte> we really need to get it written up formally at some point

[01-Sep-2011 13:12:57] <rmatte> yeh just the first rpm, that's it

[01-Sep-2011 13:13:02] <rmatte> since you won't even be running zope on it

[01-Sep-2011 13:13:10] <rmatte> and the ZenPack will take care of copying anything over that needs to be

[01-Sep-2011 13:13:15] <Thinh> yea i remembered doing this

[01-Sep-2011 13:13:21] <Thinh> gotta configure ssh using zenoss username?

[01-Sep-2011 13:13:34] <rmatte> nah, it's a bit more complex than that

[01-Sep-2011 13:14:08] <rmatte> Basically you do this: http://community.spiceworks.com/education/projects/Passwordless_SSH_Using_Shared_Keys

[01-Sep-2011 13:14:08] <hmp> dont forget the .environment file and PermitUserEnvironment Yes in sshd.conf

[01-Sep-2011 13:14:11] <rmatte> for the zenoss user

[01-Sep-2011 13:14:17] <Thinh> i know the enterprise version is painless kinda

[01-Sep-2011 13:14:30] <rmatte> then read the rest of the comments that I left on the second comments page for the zenpack

[01-Sep-2011 13:14:37] <rmatte> I have all the necessary steps listed

[01-Sep-2011 13:14:57] <hmp> Thinh: its painless here also, you just have extra steps

[01-Sep-2011 13:15:02] <Thinh> rgr that

[01-Sep-2011 13:15:46] <hmp> dont forget to edit any iptables/network ACLs along the way

[01-Sep-2011 13:15:54] <Thinh> heh yea

[01-Sep-2011 13:16:15] <Thinh> hey what does hte backup feature back up ?

[01-Sep-2011 13:16:21] <Thinh> config and templates?

[01-Sep-2011 13:16:29] <rmatte> no idea, you'd have to ask egor

[01-Sep-2011 13:16:40] <rmatte> I didn't even know there was a backup feature

[01-Sep-2011 13:16:44] <Thinh> heha

[01-Sep-2011 13:16:44] <Thinh> yea

[01-Sep-2011 13:16:51] <Thinh> well in enterprise there was unless it not there anymore

[01-Sep-2011 13:17:08] <rmatte> well, do you mean the general Zenoss backup feature, or a collector specific one?

[01-Sep-2011 13:46:28] Guest16184 is now known as Glace

[01-Sep-2011 13:46:58] Glace is now known as Guest29337

[01-Sep-2011 13:48:14] kocolosk|away is now known as kocolosk

[01-Sep-2011 13:48:36] Guest29337 is now known as Glacee

[01-Sep-2011 14:18:00] willwh`afk is now known as willwh

[01-Sep-2011 14:28:51] <kocolosk> Hi, I've got a question about Interface monitoring.  I have an eth2 and eth3 in several servers that are clearly marked DOWN in `ip a l` but zenoss continues to claim status=up and monitored=true.  How does it make that determination? (servers monitored by SNMP)

[01-Sep-2011 14:29:20] <kocolosk> iow, what would it take for an Interface to be marked down by zenoss?

[01-Sep-2011 14:30:17] <Sam-I-Am> when you remodel it should see if its up or down

[01-Sep-2011 14:30:33] <Sam-I-Am> otherwise you can do it by traps, or polling, if the device is in the correct class

[01-Sep-2011 14:30:33] <kocolosk> hmm, it's been several weeks

[01-Sep-2011 14:31:26] <kocolosk> the servers are in different device classes, one i'm looking at now is just /Server

[01-Sep-2011 14:31:44] <kocolosk> no custom config props

[01-Sep-2011 14:33:05] <Sam-I-Am> docs/DOC-2494

[01-Sep-2011 14:33:10] <Sam-I-Am> that may help, one way of doing it

[01-Sep-2011 14:33:17] <kocolosk> thanks Sam-I-Am

[01-Sep-2011 14:33:41] <Sam-I-Am> the other way is traps, which may or may not be as reliable

[01-Sep-2011 14:33:49] <kocolosk> " One thing that surprises a lot of people is that it doesn't also monitoring the up/down status of the interfaces."

[01-Sep-2011 14:33:57] <kocolosk> count me among the surprised

[01-Sep-2011 14:34:06] <kocolosk>

[01-Sep-2011 14:34:23] <Sam-I-Am> theres some weird stuff

[01-Sep-2011 14:34:28] <kocolosk> i'll probably check the snmp traps config first

[01-Sep-2011 14:34:28] <Sam-I-Am> call it configurability

[01-Sep-2011 14:34:30] <kocolosk>

[01-Sep-2011 14:35:11] <kocolosk> i don't absolutely positively need it to poll these interfaces, i just found it odd to see the little green power button when i knew for a fact those interfaces had never been brought up

[01-Sep-2011 14:36:35] <kocolosk> thanks again Sam-I-Am

[01-Sep-2011 14:36:59] <Thinh> anyone have hte doc for ldap intergration

[01-Sep-2011 14:38:50] <Thinh> are all config store in a flat file or database

[01-Sep-2011 14:39:00] <rmatte> Thinh: This is the only thing I've ever seen for 3.x: thread/14081

[01-Sep-2011 14:39:19] <rmatte> I still need to get it working on 3.x before I upgrade, if there's no guide by then I'll write one up as I go

[01-Sep-2011 14:39:32] <Thinh> damn

[01-Sep-2011 14:39:48] <rmatte> there's the old 2.xguides but they are obsolete

[01-Sep-2011 14:39:50] <Thinh> i thought you cando the ldap config thru a web portal

[01-Sep-2011 14:40:06] <rmatte> you can do it through the zope interface, yes

[01-Sep-2011 14:40:13] <Thinh> whats hte port for that

[01-Sep-2011 14:40:13] <rmatte> but installing the actual plugins is a different story

[01-Sep-2011 14:40:18] <Thinh> oh ic

[01-Sep-2011 14:40:26] <Thinh> you need a plugin to get ldap to work

[01-Sep-2011 14:40:27] <Thinh> thats bs

[01-Sep-2011 14:40:27] <rmatte> there's no port, you just add /manage after /dmd

[01-Sep-2011 14:40:41] <rmatte> blah:8080/zport/dmd/manage

[01-Sep-2011 14:40:54] <rmatte> you can add /manage at any level

[01-Sep-2011 14:41:06] <rmatte> higher levels you have to be logged in as the administrator account

[01-Sep-2011 14:41:16] <Thinh> gotcha

[01-Sep-2011 14:41:16] <rmatte> such as blah:8080/zport/manage

[01-Sep-2011 14:41:27] <Thinh> so i still need a plugin before i can configure it ?

[01-Sep-2011 14:41:34] <rmatte> you can also do it directly on lower levels to access template/device/whatever properties

[01-Sep-2011 14:41:38] <rmatte> correct

[01-Sep-2011 14:41:49] <rmatte> the plugin needs to be installed

[01-Sep-2011 14:41:56] <rmatte> It's a zope plugin

[01-Sep-2011 14:42:31] <Thinh> damn

[01-Sep-2011 14:42:58] <Thinh> i should get enterprise version but meh

[01-Sep-2011 14:43:01] <Thinh> too much $

[01-Sep-2011 14:43:13] <rmatte> It's really not that bad to install the plugins, I just haven't had a chance to work through it

[01-Sep-2011 14:43:40] <rmatte> that thread I pasted explains it fairly well

[01-Sep-2011 14:43:46] <Thinh> i have to setup a presention around october

[01-Sep-2011 14:43:50] <Thinh> trying to get everything line up

[01-Sep-2011 14:44:08] <Thinh> getting lazy now haha

[01-Sep-2011 14:44:29] <rmatte> This is the main guide: docs/DOC-2510

[01-Sep-2011 14:44:44] <rmatte> between that guide and the thread (which is linked to from the top of the guide) you should be able to sort it out easily enough

[01-Sep-2011 14:47:04] <Thinh> hmm backup stores evryrthing

[01-Sep-2011 14:47:32] <Thinh> haha where is the option to restore

[01-Sep-2011 14:51:39] <Thinh> anyone know ?

[01-Sep-2011 14:52:15] <Thinh> zenrestore! thre we go =)

[01-Sep-2011 15:10:44] <JohnnyNOC> anyone here using zenoss 3.x and rrdcache?

[01-Sep-2011 15:11:06] <JohnnyNOC> it looks like it's supposed to be in 3.1 (which i'm using) but when I look for the changes from http://dev.zenoss.com/trac/changeset/21852 they aren't there

[01-Sep-2011 15:11:07] <JohnnyNOC> :[

[01-Sep-2011 15:11:23] <Simon4> as far as I know it's only in Avalon

[01-Sep-2011 15:11:29] <Simon4> it was promised for 3 but never eventuated

[01-Sep-2011 15:11:36] <Simon4> but is apparently fairly easy to backport

[01-Sep-2011 15:12:28] <JohnnyNOC> theoretically i download the changeset and apply the patches?

[01-Sep-2011 15:12:40] <JohnnyNOC> from http://dev.zenoss.com/trac/changeset/21852

[01-Sep-2011 15:12:58] <Simon4> I'm sure there was a howto posted somewhere but I'm struggling to find it

[01-Sep-2011 15:13:25] <JohnnyNOC> i see thread/13718

[01-Sep-2011 15:15:29] <Simon4> yeah, there was another one I htought

[01-Sep-2011 15:15:36] <Simon4> but am totally and utterly failing to find it

[01-Sep-2011 15:15:46] <Simon4> was written by another Simon

[01-Sep-2011 15:18:42] <Hackman238> JohnnyNOC: Its coming in v4

[01-Sep-2011 15:19:53] <JohnnyNOC> which is like, any day now?

[01-Sep-2011 15:20:22] <Hackman238> JohnnyNOC: Look at the ZCA roadmap in the channel topic

[01-Sep-2011 15:22:08] <JohnnyNOC> h

[01-Sep-2011 15:22:12] <JohnnyNOC> it's 3.2 which is supposed to be any day now

[01-Sep-2011 15:23:07] <Hackman238> JohnnyNOC: Yep.

[01-Sep-2011 15:27:19] <Hackman238> JohnnyNOC: Let me see if my backup of the preavalon SVN has rrdcache bits

[01-Sep-2011 15:46:57] <JohnnyNOC> Hackman238 i appreciate it but no need..  i have plenty of things to keep me busy in between now and then and IO currently isn't a big deal

[01-Sep-2011 15:47:09] <JohnnyNOC> but i know we want to implement it at some point

[01-Sep-2011 15:47:39] <Hackman238> JohnnyNOC: Alrighty. Just know its coming

[01-Sep-2011 15:50:17] <JohnnyNOC> Hackman238 i have another question though

[01-Sep-2011 15:50:21] <Hackman238> JohnnyNOC: Sure

[01-Sep-2011 15:50:38] <JohnnyNOC> so i have these windows domain controllers that i can't query via WMI because of permissions, no big deal

[01-Sep-2011 15:50:47] <JohnnyNOC> i just setup a new monitoring template using a command datasource

[01-Sep-2011 15:51:02] <Hackman238> JohnnyNOC: Ok

[01-Sep-2011 15:51:15] <JohnnyNOC> ran zencommand manually but it's not creating my rrds for me

[01-Sep-2011 15:51:28] <JohnnyNOC> i was just toying around with check_dns

[01-Sep-2011 15:51:32] <JohnnyNOC> and wanted a little graph showing response times

[01-Sep-2011 15:52:01] <Hackman238> JohnnyNOC: When you run zencommand add a -v10. Zencommand debug mode will tell you the name of the rrd's it writes to

[01-Sep-2011 15:52:05] <JohnnyNOC> trying to understand why -

[01-Sep-2011 15:52:26] <JohnnyNOC> but - it doesn't

[01-Sep-2011 15:52:33] <Hackman238> JohnnyNOC: Try running it manually with -v10 to see if its working

[01-Sep-2011 15:52:37] <JohnnyNOC> let me try one more time

[01-Sep-2011 15:52:41] <JohnnyNOC> i have done that - and it appears to be yes

[01-Sep-2011 15:52:53] <Hackman238> JohnnyNOC: Does it fetch the config for the devices?

[01-Sep-2011 15:53:05] <JohnnyNOC> says it's successful and all that

[01-Sep-2011 15:53:06] <Hackman238> JohnnyNOC: If you could post the log I'll be able to help more

[01-Sep-2011 15:53:11] <JohnnyNOC> yes

[01-Sep-2011 15:53:45] <JohnnyNOC> ok

[01-Sep-2011 15:53:45] <JohnnyNOC> 1s

[01-Sep-2011 15:53:50] <JohnnyNOC> subverting firewall

[01-Sep-2011 15:55:14] <JohnnyNOC> everything seems groovy in the logs

[01-Sep-2011 15:55:23] <JohnnyNOC> but i don't see it tell me what rrd it's going to create/update

[01-Sep-2011 15:56:06] <JohnnyNOC> and it in fact, does not appear to be creating/updating one

[01-Sep-2011 15:59:23] <dhopp> is there a config file somewhere that tells zenoss whether it is initialized or not?

[01-Sep-2011 15:59:43] <dhopp> I'm working on scripting creating remote collectors and they have no need to initialize zope

[01-Sep-2011 15:59:43] <JohnnyNOC> dhopp honestly i don't know

[01-Sep-2011 15:59:54] <JohnnyNOC> but maybe you're talking about something else

[01-Sep-2011 15:59:57] <dhopp> err initilize zeodb

[01-Sep-2011 16:00:03] <rmatte> dhopp: you mean you want to stop certain daemons from loading, correct?

[01-Sep-2011 16:00:09] <dhopp> rmatte: no

[01-Sep-2011 16:00:14] <dhopp> I know how to do that

[01-Sep-2011 16:00:42] <rmatte> then you mean is there somehow that Zenoss knows if the daemon is already running?

[01-Sep-2011 16:00:43] <dhopp> but I did all of that (created the deamons.txt and DAEMONS_TXT_ONLY) but when I start zenoss on that server I get "Zenoss not initialized.  Performing first-boot initialization..."

[01-Sep-2011 16:00:56] <rmatte> oh, that's really weird

[01-Sep-2011 16:01:10] <rmatte> it might be because it finds that MySQL isn't running or something

[01-Sep-2011 16:01:27] <rmatte> I've never seen that, sounds like something new that they added in possibly

[01-Sep-2011 16:01:35] <rmatte> actually, are you running on CentOS?

[01-Sep-2011 16:01:52] <dhopp> aha!

[01-Sep-2011 16:01:53] <dhopp> # signature of zenoss having been installed for the first time (but

[01-Sep-2011 16:01:53] <dhopp> # not configured yet)

[01-Sep-2011 16:01:53] <dhopp> export FRESH_INSTALL=${ZENHOME}/.fresh_install

[01-Sep-2011 16:01:54] <rmatte> I remember I did a CentOS install a while back and it had a pretty funky startup script setup

[01-Sep-2011 16:01:58] <rmatte> ah

[01-Sep-2011 16:02:03] <rmatte> that explains it hehe

[01-Sep-2011 16:02:25] <dhopp> will need to add to my script to create that file

[01-Sep-2011 16:02:32] <rmatte> you could just let it run through, all that does is build the initial zope tables

[01-Sep-2011 16:02:43] <dhopp> yeah I know..but I don't want to wait :-)

[01-Sep-2011 16:02:50] <rmatte> I see

[01-Sep-2011 16:02:52] <dhopp> and it start mysql

[01-Sep-2011 16:02:56] <dhopp> which I don't want on that box

[01-Sep-2011 16:03:02] <rmatte> yeh

[01-Sep-2011 16:03:18] <rmatte> though it'll probably start MySQL anyways even without that

[01-Sep-2011 16:03:27] <rmatte> you need to find where in the startup scripts it starts it and comment it out

[01-Sep-2011 16:03:56] <rmatte> though on CentOS I think MySQL is actually installed separately from the Zenoss package (as it's own package)

[01-Sep-2011 16:04:15] <rmatte> so you should be able to just use the service command or whatever to disable mysql on startup

[01-Sep-2011 16:04:28] <dhopp> rmatte: I'm using RHEL and it is install separately..but zenoss is still starting it

[01-Sep-2011 16:04:57] <SDuensin> Ah HA!  I think I finally found my Windows deployment problem!   Anyone here *really* good with "winexe"?

[01-Sep-2011 16:04:58] <dhopp> yeah I will see if mysql starts after it does it's initialization (i've already turned it off on boot).  I'll stop mysql and zenoss and restart zenoss and see if mysql starts

[01-Sep-2011 16:05:00] <zuzu> hi guys

[01-Sep-2011 16:05:25] <zuzu> I have a problem with my mysql db's since I changed hostname (zenoss 3.0)

[01-Sep-2011 16:06:17] <zuzu> couldn't really find a lot of documentation on this

[01-Sep-2011 16:06:26] <zuzu> I guess i have to set my permissions right

[01-Sep-2011 16:06:36] <rmatte> zuzu: what kind of problem?

[01-Sep-2011 16:06:44] <zuzu> mysql won't start

[01-Sep-2011 16:07:01] <zuzu> ibdata is not accessible

[01-Sep-2011 16:07:05] <rmatte> you use the stack installer to install?

[01-Sep-2011 16:07:11] <zuzu> stack

[01-Sep-2011 16:07:12] <dhopp> oh actually you don't create that fresh_install file you remove it

[01-Sep-2011 16:07:13] <dhopp> heh

[01-Sep-2011 16:07:19] <rmatte> dhopp: yes

[01-Sep-2011 16:07:33] <rmatte> zuzu: make sure you have...

[01-Sep-2011 16:07:34] <rmatte> chown -R zenoss:users /usr/local/zenoss

[01-Sep-2011 16:07:34] <rmatte> chown -R root:mysql /usr/local/zenoss/mysql

[01-Sep-2011 16:07:34] <rmatte> chown -R mysql:mysql /usr/local/zenoss/mysql/data

[01-Sep-2011 16:07:34] <rmatte> chown root:zenoss /usr/local/zenoss/zenoss/bin/zensocket

[01-Sep-2011 16:07:34] <rmatte> chmod 04750 /usr/local/zenoss/zenoss/bin/zensocket

[01-Sep-2011 16:07:44] <zuzu> ha

[01-Sep-2011 16:07:47] <zuzu> will try right away

[01-Sep-2011 16:07:55] <rmatte>

[01-Sep-2011 16:10:37] <zuzu> should i restart zenoss as a whole or just the /usr/local/zenoss/zenoss/scripts/ctl.sh ?

[01-Sep-2011 16:15:58] <jmp242> SDuensin: not as such, I thought winexe didn't really work right

[01-Sep-2011 16:16:05] <rmatte> that ctl script does restart it as a whole lol

[01-Sep-2011 16:16:16] <zuzu> yeah i was wrong

[01-Sep-2011 16:16:22] <zuzu> i meant the mysql ctl.sh

[01-Sep-2011 16:16:31] <zuzu> anyway i restarted it as a whole rmatte

[01-Sep-2011 16:16:35] <rmatte> cool

[01-Sep-2011 16:16:54] <zuzu> nohup: redirecting stderr to stdout

[01-Sep-2011 16:16:59] <zuzu> Starting mysqld.bin daemon with databases from /usr/local/zenoss/mysql/data

[01-Sep-2011 16:17:04] <zuzu> STOPPING server from pid file /usr/local/zenoss/mysql/data/vps.utopit.be.pid

[01-Sep-2011 16:17:10] <zuzu> 110901 22:15:19  mysqld ended

[01-Sep-2011 16:17:34] <zuzu>  /usr/local/zenoss/mysql/scripts/ctl.sh : mysql  could not be started

[01-Sep-2011 16:18:00] <rmatte> zuzu: when you applied those permissions that I gave you you did it as root?

[01-Sep-2011 16:18:09] <zuzu> yup

[01-Sep-2011 16:18:37] <rmatte> there should be a mysql log file in /usr/local/zenoss/mysql/data/

[01-Sep-2011 16:18:42] <zuzu> correct

[01-Sep-2011 16:18:43] <rmatte> check what the actual error is

[01-Sep-2011 16:18:44] <zuzu> hang on

[01-Sep-2011 16:18:55] <zuzu> it was something with not being able to open ibdata

[01-Sep-2011 16:19:01] <SDuensin> jmp242, it kinda does if you hack on it.  But I need to do an implicit NTLM auth to a remote box.

[01-Sep-2011 16:19:22] <rmatte> can you get the whole errors and fpaste.org it?

[01-Sep-2011 16:19:48] <Hackman238> JohnnyNOC: try a rrdtool last {rrd}.rrd on one of the RRD's it should be populating

[01-Sep-2011 16:20:12] <Hackman238> SDuensin: What was it?

[01-Sep-2011 16:20:22] <SDuensin> Hackman238, lemme find a URL

[01-Sep-2011 16:20:34] <Hackman238> dhopp: Whats up?

[01-Sep-2011 16:21:14] <SDuensin> Hackman238, http://riosec.com/Windows-UAC-PsExec   (read the bit in the middle between all the screenshots about implicit  &  explicit authentication)

[01-Sep-2011 16:22:01] <zuzu> rmatte: http://pastebin.com/zxqFH3zu

[01-Sep-2011 16:23:03] <jmp242> Oh, I just disable UAC

[01-Sep-2011 16:24:34] <SDuensin> UAC is evil.

[01-Sep-2011 16:25:01] <jmp242> Though I've left it on on servers for now

[01-Sep-2011 16:25:07] <jmp242> but I'm starting to reconsider that too

[01-Sep-2011 16:25:14] <jmp242> it just causes too many things to fail

[01-Sep-2011 16:25:24] <jmp242> and there are better ways to impliment what it does IMO

[01-Sep-2011 16:26:51] <SDuensin> Yea.  It's called UNIX.

[01-Sep-2011 16:27:01] <jmp242> Well, it's also on Windows

[01-Sep-2011 16:27:11] <jmp242> I think either using something like SuRun for specific elevation

[01-Sep-2011 16:27:24] <jmp242> or use a HIPS like Comodo Internet Security

[01-Sep-2011 16:27:31] <jmp242> either works much better than UAC

[01-Sep-2011 16:27:34] <rmatte> zuzu: if you do: ps aux | grep mysql

[01-Sep-2011 16:27:38] <rmatte> do you see processes running?

[01-Sep-2011 16:27:57] <SDuensin> They just keep piling more and more crap onto Windows in an attempt to secure it.  Make it stop!

[01-Sep-2011 16:28:28] <zuzu> rmatte: hmm yes

[01-Sep-2011 16:28:31] <rmatte> stop using windows

[01-Sep-2011 16:28:31] <rmatte> lol

[01-Sep-2011 16:28:37] <zuzu> and some with my old hostname.pid

[01-Sep-2011 16:28:39] <SDuensin> I would LOVE to.

[01-Sep-2011 16:28:42] <zuzu> root of the problem

[01-Sep-2011 16:28:43] <SDuensin> Personally, I have.

[01-Sep-2011 16:28:46] <zuzu> so i'll just kill those

[01-Sep-2011 16:29:05] <rmatte> zuzu: yeh, kill them then try again, that should do it

[01-Sep-2011 16:29:26] <zuzu> okay, damn could've found that myself lol

[01-Sep-2011 16:29:31] <zuzu> sorry for that

[01-Sep-2011 16:29:45] <rmatte> no problem

[01-Sep-2011 16:30:21] <jmp242> Ok ... work is over - see you all tomorrow.

[01-Sep-2011 16:31:53] <zuzu> rmatte: so restart /usr/local/zenoss/mysql/scripts/ctl.sh should only restart the db service

[01-Sep-2011 16:32:02] <zuzu> or should I better restart everything

[01-Sep-2011 16:33:48] <zuzu> hmm zenping and zentrap still don't seem to be able to run

[01-Sep-2011 16:35:35] <SDuensin> Anyone know anything about SAMBA4?

[01-Sep-2011 16:35:44] <Thinh> great dance

[01-Sep-2011 16:35:56] <SDuensin> Yea, I liked the Dreamcast game, too.

[01-Sep-2011 16:36:44] <Thinh> whats a dreamcast

[01-Sep-2011 16:36:51] <SDuensin> !!!!

[01-Sep-2011 16:36:56] <Thinh> jk =)

[01-Sep-2011 16:37:02] <SDuensin> Whew.  :-P

[01-Sep-2011 16:37:07] <Thinh> i have 2

[01-Sep-2011 16:37:07] <Thinh> heh

[01-Sep-2011 16:37:20] <Thinh> mvc still fun

[01-Sep-2011 16:37:45] <SDuensin> I *really* need a tool like "cmdkey.exe" from Windows for my Zenoss machine.

[01-Sep-2011 16:39:09] <Hackman238> SDuensin: Madness.

[01-Sep-2011 16:39:47] <SDuensin> Hackman238, well, I need to be authenticated before I connect so I get an implicit login that will give me the higher token.   Arrrrrrrrgh!

[01-Sep-2011 16:41:40] <Hackman238> SDuensin: Tricky situation

[01-Sep-2011 16:41:56] <SDuensin> I want to do it without needing a Windows "bounce box" in the middle.

[01-Sep-2011 16:41:59] <Hackman238> jmp242: Later man

[01-Sep-2011 16:43:58] <Thinh> where are the monitoring configs store in zenoss

[01-Sep-2011 16:50:21] <rmatte> Thinh: what monitoring configs are you talking about?

[01-Sep-2011 16:50:30] <rmatte> there are properties and there are templates

[01-Sep-2011 16:50:50] <Thinh> the properties

[01-Sep-2011 16:51:00] <Thinh> well iguess both

[01-Sep-2011 16:51:18] <rmatte> you should read the Zenoss Admin Guide, it explains all of this

[01-Sep-2011 16:51:38] <rmatte> http://downloads.sourceforge.net/zenoss/Zenoss_Administration_06-102010-3.0-v03.pdf

[01-Sep-2011 16:51:42] <rmatte> It's well written

[01-Sep-2011 16:51:56] <rmatte> read it from top to bottom and it'll answer 90% of the questions that you're likely to ask in here

[01-Sep-2011 16:52:23] <Thinh> yea true

[01-Sep-2011 16:52:52] <Thinh> RTFM

[01-Sep-2011 16:52:59] <rmatte> basically

[01-Sep-2011 16:53:16] <rmatte> I read through that in it's entirety when I first started using Zenoss and it was really helpful

[01-Sep-2011 16:53:58] <Thinh> yea i did too but i switched to opennms here and forgot everything

[01-Sep-2011 16:54:12] <rmatte> well have another read at it lol

[01-Sep-2011 16:54:34] <rmatte> If you knew it before you'll probably just have to skim over it

[01-Sep-2011 16:55:05] <Thinh> it was alot easier consulting the oracle

[01-Sep-2011 16:55:05] <Thinh> har

[01-Sep-2011 16:55:44] <rmatte> lol

[01-Sep-2011 16:56:09] <rmatte> I follow the teach a man to fish mentality with a lot of the more basic Zenoss stuff

[01-Sep-2011 16:56:37] <Thinh> just outsource the job

[01-Sep-2011 16:57:12] <klone> i like the 'build a man a fire' mentality better than 'teach a man to fish'

[01-Sep-2011 17:04:25] <rmatte> lol

[01-Sep-2011 17:20:08] <dhopp> I'm trying to add a ldap user (who has already logged in) to one of the roles and I get the error 'user blah not found'

[01-Sep-2011 17:20:19] <dhopp> but I can login as that user without problems

[01-Sep-2011 17:24:22] <rmatte> dhopp: you don't add users directly to roles, you create groups, then assign roles to groups and add the users to those groups

[01-Sep-2011 17:24:25] <rmatte> that's the proper way to do it

[01-Sep-2011 17:24:37] <dhopp> rmatte: do they have to be ldap groups?

[01-Sep-2011 17:24:38] <rmatte> (you do this in ZMI)

[01-Sep-2011 17:24:42] <rmatte> yes, they do

[01-Sep-2011 17:25:39] <rmatte> with ldap it's all dynamic upon login as far as the configuration on the Zenoss end goes

[01-Sep-2011 17:26:02] <rmatte> so you can set a role on a user locally, but when ldap logs that user in Zenoss will set it right back to the default role

[01-Sep-2011 17:26:19] <rmatte> which is why you need to use the groups

[01-Sep-2011 17:27:54] <dhopp> ok..I'll do it that way.  I had the LDAP groups working but for reasons I don't want to get into we were going to use ldap for authentication and local groups/roles for authorization

[01-Sep-2011 17:28:12] <rmatte> yeh, that's not going to work lol

[01-Sep-2011 17:33:00] <nyeates> Where might I find the code that sets the dedup event field? /Products/ZenEvents ? /Products/ZenCollector?

[01-Sep-2011 17:33:15] <nyeates> i need a starting point to go looking

[01-Sep-2011 17:33:31] <Simon4> ZenEvents from memory

[01-Sep-2011 17:33:40] <Simon4> I'm sure I've been in there before

[01-Sep-2011 17:33:45] <rmatte> cd $ZENHOME/Products  &&  grep -ir "dedupid" *

[01-Sep-2011 17:33:47] <rmatte>

[01-Sep-2011 17:33:48] <nyeates> cool, that is where i thought too...grepping now

[01-Sep-2011 17:51:37] kocolosk is now known as kocolosk|away

[01-Sep-2011 18:13:16] kocolosk|away is now known as kocolosk

[01-Sep-2011 18:14:19] <zuzu> rmatte: you still there?

[01-Sep-2011 18:17:41] <zuzu> okay nvm rmatte, fyi this helped: message/60682

[01-Sep-2011 18:30:33] <rmatte> cool

[01-Sep-2011 18:32:19] <nyeates> dedupfields = [evt.device, evt.component, evt.eventClass, evt.eventKey, evt.severity]

[01-Sep-2011 18:32:23] <nyeates> mydedupid = '|'.join(map(str, dedupfields))

[01-Sep-2011 18:32:48] <nyeates> the str() is a built-in python function.... but it is stripping the special characters out of the variables in dedupfields

[01-Sep-2011 18:33:08] <nyeates> anyone know how i can have it not strip special characters like \n

[01-Sep-2011 18:35:06] <rmatte> nyeates: it shouldn't be stripping them... and it's not even using str()

[01-Sep-2011 18:35:14] <rmatte> it looks like the "str" in that is a variable

[01-Sep-2011 18:36:16] <nyeates> the map() function is a special function that says "ok, go apply this function (str in our case) to each of the following iteratives"

[01-Sep-2011 18:36:41] <rmatte> ah I see

[01-Sep-2011 18:37:59] <nyeates> I tried mydedupid = '|'.join(dedupfields)

[01-Sep-2011 18:38:18] <nyeates> but some of the fields are int, and so threw errors

[01-Sep-2011 18:38:37] <nyeates> how do you type cast in python?

[01-Sep-2011 18:38:45] <nyeates> str(dedupfields)

[01-Sep-2011 18:38:45] <nyeates> ?

[01-Sep-2011 18:39:05] <rmatte> >>> blah = "blah\nblah"

[01-Sep-2011 18:39:05] <rmatte> >>> print blah

[01-Sep-2011 18:39:05] <rmatte> blah

[01-Sep-2011 18:39:05] <rmatte> blah

[01-Sep-2011 18:39:05] <rmatte> >>> str(blah)

[01-Sep-2011 18:39:05] <rmatte> 'blah\nblah'

[01-Sep-2011 18:39:05] <rmatte> >>> print str(blah)

[01-Sep-2011 18:39:06] <rmatte> blah

[01-Sep-2011 18:39:06] <rmatte> blah

[01-Sep-2011 18:39:11] <rmatte> definitely not the str function that's doing it

[01-Sep-2011 18:39:50] <nyeates> i think u helped me!

[01-Sep-2011 18:39:58] <rmatte> oh?

[01-Sep-2011 18:40:19] <nyeates> i want that format like 'blah\nblah'

[01-Sep-2011 18:40:23] <rmatte> ah

[01-Sep-2011 18:40:30] <nyeates> ill try that

[01-Sep-2011 18:40:33] <rmatte> print will interpret it literally

[01-Sep-2011 18:40:44] <rmatte> you won't be able to print it by just putting that on a line though

[01-Sep-2011 18:40:52] <rmatte> that only works at the python console as far as I'm aware

[01-Sep-2011 18:41:26] <nyeates> so the real problem here is..... if you look at events that have \n in the summary or elsewhere.....they are literally shown as blah\n in the summary of the event

[01-Sep-2011 18:41:43] <nyeates> i looked in the code, and we never implemented the code that was suppoed to strip that crap

[01-Sep-2011 18:41:49] <rmatte> ah

[01-Sep-2011 18:42:00] <nyeates> so now i gotta work around it in a transform

[01-Sep-2011 18:42:31] <Thinh> hey is there any doc on web transaction monitoring ?

[01-Sep-2011 18:42:37] <rmatte> \n works in messages but not in the summary

[01-Sep-2011 18:42:41] <rmatte> I know that for a fact

[01-Sep-2011 18:42:56] <rmatte> as in evt.message

[01-Sep-2011 18:43:05] <rmatte> what are you trying to do exactly?

[01-Sep-2011 18:43:22] <rmatte> just clean up dedupids?

[01-Sep-2011 18:44:43] <nyeates> no, a transform that does stuff based on the events count

[01-Sep-2011 18:45:07] <nyeates> evt.count does not exist tho, so we have to manually build a dedupid to look it up in the mysql

[01-Sep-2011 18:45:16] <rmatte> I've tried that before and it just doesn't work

[01-Sep-2011 18:45:37] <rmatte> I think it's because it's because the current event count doesn't get calculated until after the transform is processed

[01-Sep-2011 18:45:58] <rmatte> oops, accidentally type "it's because" twice there lol

[01-Sep-2011 18:46:19] <rmatte> as far as I can tell, as far as a transform is concerned the event count is always 1

[01-Sep-2011 18:47:58] <rmatte> I mean, you can play tricks and have the transform itself check for events matching the dedupid and see what the count of those are using a few of the event functions

[01-Sep-2011 18:48:09] <nyeates> it works tho so far, i have a post on it.....it works on events where summary is simple

[01-Sep-2011 18:48:21] <rmatte> ah

[01-Sep-2011 18:48:26] <nyeates> yeah i do what u said above

[01-Sep-2011 18:49:02] <rmatte> it'd be nice if the count were actually usable without doing that nonsense

[01-Sep-2011 18:49:27] <nyeates> yeah im not so sure why its not....even if it WAS the old count

[01-Sep-2011 18:50:04] <rmatte> well think of it this way...

[01-Sep-2011 18:50:44] <nyeates> i think they were afraid to expose it because the existing event has not yet gone through various workflow that could change the count

[01-Sep-2011 18:51:01] <rmatte> single events comes in (as though it were a brand new event with a count of 1) -> transform is run against the event -> deduplication occurs after the transform has been run and the count and last time values of an existing event are increased

[01-Sep-2011 18:51:23] <nyeates> yea

[01-Sep-2011 18:51:46] <rmatte> I guess logically there's no other way

[01-Sep-2011 18:52:39] <nyeates> pray is fixed in the new avalon code ;-)

[01-Sep-2011 18:52:40] <rmatte> one feature that Zenoss lacks which is killing me is the ability to set priority escalation of events at a certain count

[01-Sep-2011 18:52:49] <rmatte> there should be an option on event mappings to do that

[01-Sep-2011 18:52:50] <Thinh> what you mean

[01-Sep-2011 18:52:52] <Thinh> oh

[01-Sep-2011 18:53:01] <Thinh> you lucky there is even a count =0

[01-Sep-2011 18:53:06] <rmatte> event starts off as information, when count hits 3, make it critical

[01-Sep-2011 18:53:07] <rmatte> or whatever

[01-Sep-2011 18:53:08] <nyeates> im sure they overhauled this code in the new avalon / 4.0 stuff that will be coming to core

[01-Sep-2011 18:53:09] <Thinh> i really hate the graphs

[01-Sep-2011 18:53:24] <Thinh> why cant i select a date to view instead of using that scroller

[01-Sep-2011 18:53:26] <rmatte> right now I'm having to do that sort of thing via code that I added to our ticket creation script

[01-Sep-2011 18:53:44] <rmatte> nyeates: yeh, can't wait to see the new stuff

[01-Sep-2011 18:53:53] <Thinh> why dont oyu make a rule set that say after 3 count alert =0

[01-Sep-2011 18:54:11] <Thinh> does it have to be label critical

[01-Sep-2011 18:54:39] <rmatte> I don't use the standard alerting rules/filters that most people use

[01-Sep-2011 18:54:48] <nyeates> :-)

[01-Sep-2011 18:54:52] <rmatte> I use a custom daemon that generates tickets in our ticketing system

[01-Sep-2011 18:55:06] <rmatte> and it generates no matter what if they are a certain severity or higher

[01-Sep-2011 18:55:26] <rmatte> but if I could start events at lower severities and then move them up when they hit 3 I'd have more control

[01-Sep-2011 19:08:05] <tvincent_> rmatte: there is a page on the wiki for that.

[01-Sep-2011 19:13:57] <rmatte> tvincent_: probably using a transform though which I'm not interested in

[01-Sep-2011 19:14:38] <tvincent_> yea it was.

[01-Sep-2011 19:15:11] <tvincent_> you could monitor the event count. not sure how your daemon watch's the events table

[01-Sep-2011 19:15:36] <rmatte> http://dmon.org/zenticket.html

[01-Sep-2011 19:15:46] <rmatte> There's the details of the daemon including the code (at the bottom)

[01-Sep-2011 19:15:56] <rmatte> the code is sloppy since I was just learning python at the time

[01-Sep-2011 19:16:51] <tvincent_> lol

[01-Sep-2011 19:16:53] <rmatte> I will update it eventually

[01-Sep-2011 19:16:55] <tvincent_> hey it works

[01-Sep-2011 19:17:02] <rmatte> yeh, it works quite well actually lol

[01-Sep-2011 19:17:09] <rmatte> on the surface it looks polished

[01-Sep-2011 19:17:48] <tvincent_> at the end of the day....

[01-Sep-2011 19:18:05] <rmatte> I'm generally a stickler for clean code though

[01-Sep-2011 19:18:26] <rmatte> the code in that isn't really that bad, it just needs to be procedure oriented

[01-Sep-2011 19:18:28] <otakup0pe> (at the end of the day you're another day older)

[01-Sep-2011 19:18:45] <rmatte> alright, I've gotta go, later guys

[01-Sep-2011 19:46:14] willwh is now known as willwh`afk

[01-Sep-2011 22:48:58] kocolosk is now known as kocolosk|away

[02-Sep-2011 00:00:01] [disconnected at Fri Sep  2 00:00:01 2011]

[02-Sep-2011 00:00:02] [connected at Fri Sep  2 00:00:02 2011]

[02-Sep-2011 00:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[02-Sep-2011 02:31:47] <konvert> hi guys i have a question

[02-Sep-2011 02:32:16] <Thinh> its pretty late =0

[02-Sep-2011 02:38:21] <konvert> here it is 8.30 in the morning

[02-Sep-2011 02:58:19] <konvert> anybody on?

[02-Sep-2011 03:14:33] <hmp> aye

[02-Sep-2011 03:20:20] <konvert> i have a question about mibs

[02-Sep-2011 05:54:23] <luizzmizz> hi there

[02-Sep-2011 05:56:14] <luizzmizz> i'm writing a modeler and i wanted a little debug... I use a logger = logging.getLogger('zen.ZenModeler') and put a logger.debug('HI!') at the beggining of the process... it should work like that, isn't it?

[02-Sep-2011 05:57:49] <luizzmizz> (I should mind time at US is not the same than europe when i'm asking :S:S:S)

[02-Sep-2011 06:17:10] <luizzmizz> well if anyone read my question at 4am... i believe the answer is plugin.condition

[02-Sep-2011 07:40:36] <konvert> hello

[02-Sep-2011 07:41:04] <konvert> somebody who can help me with mibs

[02-Sep-2011 07:59:29] kocolosk|away is now known as kocolosk

[02-Sep-2011 08:01:00] <photon01> Hi everyone

[02-Sep-2011 08:01:32] <photon01> I need some help with an SNMP datasource

[02-Sep-2011 08:03:38] <photon01> some1 can help plz ,

[02-Sep-2011 08:03:38] <photon01> ?

[02-Sep-2011 08:04:17] <konvert> photon01 i think it's still to early in US

[02-Sep-2011 08:05:01] <photon01> I guess so

[02-Sep-2011 08:05:05] <photon01> thanks anyway

[02-Sep-2011 08:30:04] <jmp242> morning all

[02-Sep-2011 08:30:16] <photon01> morning

[02-Sep-2011 08:30:56] <jmp242> hey, what's your problem with SNMP datasource?

[02-Sep-2011 08:31:07] <photon01> yeah, was going to ask u

[02-Sep-2011 08:31:09] <photon01> well

[02-Sep-2011 08:31:16] <photon01> I added a template

[02-Sep-2011 08:31:21] <photon01> to a router

[02-Sep-2011 08:31:26] <photon01> local template

[02-Sep-2011 08:31:42] <photon01> then added a datasource for an interface speed (rrd type derive)

[02-Sep-2011 08:31:58] <photon01> I tested it, it tests well, so no snmp issue

[02-Sep-2011 08:32:02] <photon01> but then no graph

[02-Sep-2011 08:32:10] <photon01> and no rrd file created

[02-Sep-2011 08:32:24] <photon01> (i mean only NaNs in graphs I created)

[02-Sep-2011 08:32:48] <jmp242> interfaces generally use component templates

[02-Sep-2011 08:32:54] <photon01> yes

[02-Sep-2011 08:32:58] <jmp242> are you trying to create a perf template for this?

[02-Sep-2011 08:33:04] <photon01> yes

[02-Sep-2011 08:33:14] <photon01> well the component works well

[02-Sep-2011 08:33:27] <photon01> but I want to sum 2 interface speeds in a graph

[02-Sep-2011 08:33:31] <jmp242> are you sure the OID you're grabbing isn't already being collected by Zenoss?

[02-Sep-2011 08:33:35] <photon01> that's why I created the template

[02-Sep-2011 08:33:42] <photon01> it probably is

[02-Sep-2011 08:33:46] <jmp242> that's the problem

[02-Sep-2011 08:33:51] <photon01> okaaay

[02-Sep-2011 08:33:53] <jmp242> Zenoss doesn't let you use the same OID twice

[02-Sep-2011 08:33:54] <photon01> didn"t know that

[02-Sep-2011 08:33:58] <jmp242> we think it's a bug

[02-Sep-2011 08:34:00] <photon01> okay

[02-Sep-2011 08:34:05] <jmp242> Zenoss seems to thing it's a design decision

[02-Sep-2011 08:34:08] <photon01> now that explains it

[02-Sep-2011 08:34:11] <jmp242> so . . .

[02-Sep-2011 08:34:27] <photon01> so the only solution is to use a command datasource

[02-Sep-2011 08:34:30] <photon01> with snmpwalk

[02-Sep-2011 08:34:31] <jmp242> I think so

[02-Sep-2011 08:34:47] <photon01> that does clear things up !

[02-Sep-2011 08:34:49] <photon01> thanks alot

[02-Sep-2011 08:34:52] <jmp242> glad to help

[02-Sep-2011 08:34:55] <Tabmow> Anyone successfully monitoring JBoss from zenoss 3? We have the enterprise zenpacks, I have followed the documentation but am getting a bunch of error events which for the life of me I can't figure out

[02-Sep-2011 08:34:56] <photon01>

[02-Sep-2011 08:35:34] <jmp242> sorry Tabmow I have neither JBoss nor Enterprise

[02-Sep-2011 08:35:49] <jmp242> If I had enterprise and that problem, I'd probably open a support ticket

[02-Sep-2011 08:36:35] <Tabmow> I might do that... I believe the problem is on the JBoss end most likely but I will look into it

[02-Sep-2011 08:37:58] <konvert> jmp242: i'm stil struggling with the mibs and traps translations.

[02-Sep-2011 08:38:31] <Hackman238> Morning all

[02-Sep-2011 08:39:37] <jmp242> Hey Hackman238

[02-Sep-2011 08:39:49] <jmp242> Do you know if zenbatchload supports multiple groups

[02-Sep-2011 08:40:02] <jmp242> at a time with setGroups

[02-Sep-2011 08:40:29] <jmp242> I looked at the code myself, and it's not clear, but it really does look like it's one group at a time

[02-Sep-2011 08:40:55] <konvert> Hackman238 and jmp242 i have posted an post in the forum about my mibs

[02-Sep-2011 08:41:06] <jmp242> message/61004#61004

[02-Sep-2011 08:41:09] <jmp242> is it that one?

[02-Sep-2011 08:41:11] <jmp242> it's on my list

[02-Sep-2011 08:42:22] <klone> jmp242, yes ... you should be able to associate multiple groups via zenbatchload

[02-Sep-2011 08:42:25] <jmp242> konvert: I will say that MIBS seem to be a bit of a flaky part of Zenoss, especially as they're not heavily used - just for incoming traps, and it's often easier to work around with event transforms than trying to get a good MIB file and Zenoss to agree

[02-Sep-2011 08:42:51] <jmp242> klone - want to post an answer to this thread? message/60984#60984

[02-Sep-2011 08:42:58] <jmp242> it's been hanging around since last year

[02-Sep-2011 08:46:34] <konvert> yes jmp242 it's that one

[02-Sep-2011 08:47:01] <konvert> so i should stop messing up with mibs and translate the traps myself?

[02-Sep-2011 08:47:41] <jmp242> Well, it depends on you I guess more than anything

[02-Sep-2011 08:47:57] <jmp242> if you do the translations yourself, you'll be getting some working as you go

[02-Sep-2011 08:48:17] <jmp242> if you mess with the MIBs, you might get nothing, you might get everything working at some point in time

[02-Sep-2011 08:48:38] <jmp242> I don't mess with MIBs myself, but I don't do much with Traps so . . .

[02-Sep-2011 08:48:54] <Hackman238> konvert: Alrighty

[02-Sep-2011 08:53:20] <photon01> I have about two unrelated questions

[02-Sep-2011 08:54:06] <photon01> 1) When I implement an SNMP collection using a command datasource, do I need to parse the output or does the Auto Parser understands it ?

[02-Sep-2011 08:54:24] <Hackman238> konvert: And those traps are indeed in those mibs?

[02-Sep-2011 08:55:04] <konvert> yes they are

[02-Sep-2011 08:55:13] <photon01> 2) Do i need to install all zenpacks before doing a restore ?

[02-Sep-2011 08:56:03] <Hackman238> photon01: stack or native?

[02-Sep-2011 08:56:17] <konvert> Hackman238 yes they are

[02-Sep-2011 08:56:24] <Hackman238> konvert: Gotcha

[02-Sep-2011 08:56:59] <photon01> <Hackman238> : From stack to stack

[02-Sep-2011 08:57:18] <photon01> <Hackman23> : Same version (3.1) and mysql data included

[02-Sep-2011 08:57:55] <Hackman238> photon01: Using zenrestore?

[02-Sep-2011 08:58:28] <photon01> <Hackman23> : yes. Actually I'm testing backups, if there's a better way, I take it

[02-Sep-2011 08:58:47] <jmp242> photon01: for SNMP collection using command datasource, you do need to convert the output to nagios style yourself in your script

[02-Sep-2011 08:59:18] <Hackman238> photon01: I'd reinstall the packs before you use zenrestore only because its been known to not restore zenpacks right

[02-Sep-2011 08:59:25] <photon01> <jmp24> : Okay, thanks again. I think i'm going to use sed to keep it light

[02-Sep-2011 08:59:33] <Hackman238> photon01: NP

[02-Sep-2011 09:00:18] <Hackman238> photon01: Whoops I just realized you were talking to jmp242

[02-Sep-2011 09:00:22] <Hackman238>

[02-Sep-2011 09:00:43] <jmp242> NOT OFFICIAL: I still like not using zenbackup and instead just rsyncing the $ZENHOME and doing a MySQL Dump

[02-Sep-2011 09:01:02] <photon01> <Hackman238> : lol. okay well thank you too

[02-Sep-2011 09:01:45] <photon01> <jmp242> : okay. I saw it in forum. Well, once I test it I'm going to come back

[02-Sep-2011 09:13:37] <konvert> hi guys someone has the article for transforming traps for zenoss 3.0+ because i only find the old article http://docs.huihoo.com/zenoss/admin-guide/2.1.1/ch10s15.html

[02-Sep-2011 09:18:37] <jmp242> docs/DOC-9437#SNMPEVENTTransform

[02-Sep-2011 09:18:47] <jmp242> It's basically the same except perhaps the Web GUI

[02-Sep-2011 09:19:51] <konvert> thx

[02-Sep-2011 09:25:44] <konvert> looking into it atm

[02-Sep-2011 09:38:09] <SDuensin> FrIdAy!

[02-Sep-2011 09:38:19] <dec3pti0n> it seems like me trying to move a report to another organizer caused some zenoss db issue I think .. anyway to correct this "KeyError: 'Apache Requests'" error

[02-Sep-2011 09:55:02] <dec3pti0n> I tried this message/55666 and also trying to reindex the zodb

[02-Sep-2011 09:59:14] <JohnnyNOC> du0d

[02-Sep-2011 09:59:27] <JohnnyNOC> why does zenoss have a problem with production states and maintenance windows in 3.1.0

[02-Sep-2011 09:59:40] <JohnnyNOC> i have a bunch of staging/dev machines that when a maintenance window is completed, go back to being in production

[02-Sep-2011 09:59:45] <JohnnyNOC> :[

[02-Sep-2011 10:00:11] <JohnnyNOC> stop production state = 'original'

[02-Sep-2011 10:00:31] <JohnnyNOC> so why does original == production when they're in staging/dev before the maintenance window?

[02-Sep-2011 10:01:39] <JohnnyNOC> am I going to have to create multiple maintenance windows?

[02-Sep-2011 10:02:35] <jmp242> JohnnyNOC: probably - I know that Original doesn't always work right

[02-Sep-2011 10:03:04] <jmp242> There were lots of bug reports back a while ago, and I don't know they ever really got resolved

[02-Sep-2011 10:03:30] <JohnnyNOC> ok

[02-Sep-2011 10:03:41] <JohnnyNOC> well

[02-Sep-2011 10:04:07] <JohnnyNOC> f that

[02-Sep-2011 10:04:49] <dec3pti0n> is there some tool to read the zodb  ? something besides the zendmd

[02-Sep-2011 10:05:30] <dec3pti0n> this KeyError I'm getting is just stupid and should be a quick fix on a db

[02-Sep-2011 10:06:01] <JohnnyNOC> yes

[02-Sep-2011 10:06:10] <JohnnyNOC> try using the zope management interface

[02-Sep-2011 10:06:21] <JohnnyNOC> or write a script to dump whtever you need out of zendmd

[02-Sep-2011 10:06:23] <JohnnyNOC> but whatever you do

[02-Sep-2011 10:06:35] <JohnnyNOC> don't try and set a maintenance window and expect original production states to work

[02-Sep-2011 10:07:17] <JohnnyNOC> and have alerting rules that depend on production states

[02-Sep-2011 10:07:46] <jmp242> Do you have overlapping maintenance windows?

[02-Sep-2011 10:07:50] <jmp242> that's usually where it breaks

[02-Sep-2011 10:07:52] <JohnnyNOC> i have -1-

[02-Sep-2011 10:08:02] <JohnnyNOC> and even that didn't freaking work right

[02-Sep-2011 10:08:07] <dec3pti0n> ok so I got install this "zope management interface" or comes with zenoss ?

[02-Sep-2011 10:08:09] <JohnnyNOC> it's for the windows boxes that get bounced at midnight

[02-Sep-2011 10:08:20] <jmp242> hmmm, definitely sounds like a bug

[02-Sep-2011 10:08:35] <JohnnyNOC> i can't apply it to /Devices/CIM and have it work

[02-Sep-2011 10:08:36] <jmp242> I'd open a track ticket . . . or see if there's an existing one you can add to

[02-Sep-2011 10:08:39] <JohnnyNOC> i had to apply it to /Devices/CIM/WMI

[02-Sep-2011 10:08:44] <JohnnyNOC> i'm a core user

[02-Sep-2011 10:08:51] <JohnnyNOC> i don't believe i can open a ticket

[02-Sep-2011 10:08:54] <jmp242> yea, trac is for core

[02-Sep-2011 10:08:55] <jmp242> see

[02-Sep-2011 10:09:01] <jmp242> http://dev.zenoss.com/trac/wiki/HowToAddTicket

[02-Sep-2011 10:09:10] <jmp242> core users can open a bug report

[02-Sep-2011 10:09:20] <jmp242> but not support requests

[02-Sep-2011 10:09:54] <JohnnyNOC> i'd rather just bitch about to be perfectly honest

[02-Sep-2011 10:10:01] <jmp242> oh ok

[02-Sep-2011 10:10:06] <jmp242> if it makes you feel better

[02-Sep-2011 10:10:41] <jmp242> I'm going to go out on a limb, without checking so this may change later

[02-Sep-2011 10:10:57] <jmp242> but if it's something like this that you care about, Core users could post a discussion in the ZCA space

[02-Sep-2011 10:11:12] <jmp242> that way the ZCA can add it to the list for them to work on as we ramp up

[02-Sep-2011 10:11:31] <dhopp> JohnnyNOC: so if you have maintenance window that changes the state from "production" to "maintenance" are you saying that it may not revert back?  In my testing it seems to work

[02-Sep-2011 10:11:47] <JohnnyNOC> i'm saying i have devices that are in production,d evelopment, and staging

[02-Sep-2011 10:12:04] <JohnnyNOC> when my maintenance window kicks off it puts them in maintenance, and apparently a lot of these development and staging servers get put back into production

[02-Sep-2011 10:12:22] <JohnnyNOC> and my alerts i have setup which use the production state as a rule start firing off alerts that shouldn't be sent

[02-Sep-2011 10:12:25] <JohnnyNOC> and i'm the asshole

[02-Sep-2011 10:12:31] <dhopp> JohnnyNOC: and you have it set to "original state" in the mainteance?

[02-Sep-2011 10:12:38] <JohnnyNOC> yes

[02-Sep-2011 10:12:41] <dhopp> hrm

[02-Sep-2011 10:12:43] <dhopp> interesting

[02-Sep-2011 10:12:49] <JohnnyNOC> it's the ONLY mainenance window on this server

[02-Sep-2011 10:12:56] <JohnnyNOC> and it wouldn't work when i applied to /Devices/CIM

[02-Sep-2011 10:13:04] <JohnnyNOC> (using bigegors wmi datasource/perf zenpacks)

[02-Sep-2011 10:13:07] <dhopp> I'll have to try that…I guess I only worried about going from production to maint back to production

[02-Sep-2011 10:13:10] <JohnnyNOC> i ahd to apply it to /Devices/CIM/WMI

[02-Sep-2011 10:13:15] <konvert> how to clear the history of events?

[02-Sep-2011 10:13:43] <dhopp> JohnnyNOC: oh…I was using Groups to set the maintenance windows and not setting it on the device class

[02-Sep-2011 10:13:55] <JohnnyNOC> i'm talking about production states

[02-Sep-2011 10:14:03] <dhopp> JohnnyNOC: yes I know

[02-Sep-2011 10:14:13] <JohnnyNOC> oh oh

[02-Sep-2011 10:14:15] <JohnnyNOC> i get ya

[02-Sep-2011 10:14:33] <dhopp> but I'm saying I didn't set the window on a device class or on an individual device.  I put the devices into a group and set the window on that group...

[02-Sep-2011 10:14:35] <Hackman238> Sorry folks, I have a full plate today.

[02-Sep-2011 10:14:47] <dhopp> but I only worried about devices that were already production

[02-Sep-2011 10:15:08] <JohnnyNOC> well

[02-Sep-2011 10:15:12] <JohnnyNOC> i don't know how you use groups but that wouldn't work for us

[02-Sep-2011 10:15:34] <JohnnyNOC> guess it doesn't really make a difference hto

[02-Sep-2011 10:15:56] <JohnnyNOC> but in this case, doesn't make sense to create a windows group or system when we already have the device class

[02-Sep-2011 10:16:14] <jmp242> Yea, I haven't ever tested going from pre-production to maintenance and back

[02-Sep-2011 10:16:36] <jmp242> let me see

[02-Sep-2011 10:16:52] <JohnnyNOC> i'm using 3.1.0

[02-Sep-2011 10:16:56] <JohnnyNOC> just fyi

[02-Sep-2011 10:17:14] <jmp242> I'm also on 3.1.0

[02-Sep-2011 10:17:28] <JohnnyNOC> well

[02-Sep-2011 10:17:34] <JohnnyNOC> for the sake of you not thinking i'm a whack job, i hope it fails.

[02-Sep-2011 10:17:35] <dhopp> JohnnyNOC: we have a patch schedule where we do patches to dev one week, then next week patch evens (in a multi server from we label them 01,02,03,04) and then the next week patch odds.  I created those "Patch Groups"

[02-Sep-2011 10:18:43] <jmp242> This may take a little while to test

[02-Sep-2011 10:19:16] <konvert> is there a way to delete the whole history list?

[02-Sep-2011 10:19:23] <JohnnyNOC> konvert yes

[02-Sep-2011 10:19:25] <dhopp> JohnnyNOC: but I never actually tested a state that wasn't already production

[02-Sep-2011 10:20:35] <JohnnyNOC> well i'm telling you from experience

[02-Sep-2011 10:20:37] <JohnnyNOC> it's not working

[02-Sep-2011 10:20:37] <konvert> JohnnyNOC and how to do it?

[02-Sep-2011 10:21:31] <JohnnyNOC> konvert there are a bunch fo different ways

[02-Sep-2011 10:21:39] <JohnnyNOC> why don't you tell us what the problem is you're trying to solve

[02-Sep-2011 10:21:44] <JohnnyNOC> you just want to delete the events?

[02-Sep-2011 10:21:51] <JohnnyNOC> from the dashboard?

[02-Sep-2011 10:21:53] <JohnnyNOC> from the db?

[02-Sep-2011 10:21:58] <JohnnyNOC> is your db taking up too much disk space?

[02-Sep-2011 10:22:02] <JohnnyNOC> need more data

[02-Sep-2011 10:22:04] <konvert> yeah it's very slow loading, drom the dashboard

[02-Sep-2011 10:22:23] <JohnnyNOC> you should look into zenoss runing

[02-Sep-2011 10:22:23] <JohnnyNOC> er

[02-Sep-2011 10:22:24] <JohnnyNOC> tuning

[02-Sep-2011 10:22:33] <JohnnyNOC> you can tune mysql, you can tune the zope db, you can tune your OS

[02-Sep-2011 10:22:49] <JohnnyNOC> that should help the dashboard be more responsive

[02-Sep-2011 10:23:41] <dec3pti0n> thanks zmi helped me fix the issue

[02-Sep-2011 10:24:07] <JohnnyNOC> dec3pti0n you're welcome

[02-Sep-2011 10:29:03] <jmp242> JohnnyNOC: Confirmed

[02-Sep-2011 10:29:11] <jmp242> I'll look for a trac report

[02-Sep-2011 10:29:56] <JohnnyNOC> thanks jmp242

[02-Sep-2011 10:31:39] <JohnnyNOC> if you're bored you could confirm another what looks like a bug

[02-Sep-2011 10:31:42] <JohnnyNOC>

[02-Sep-2011 10:31:58] <JohnnyNOC> i have networkDisk in zFileSystemMapIgnoreTypes for /Devices/CIM/WMI devices

[02-Sep-2011 10:31:59] <jmp242> http://dev.zenoss.com/trac/ticket/7751

[02-Sep-2011 10:32:02] <JohnnyNOC> and well, they aren't being ignored

[02-Sep-2011 10:32:05] <jmp242> Going to be fixed in 3.2

[02-Sep-2011 10:32:11] <jmp242> but there are instructions there to fix in 3.1

[02-Sep-2011 10:32:51] <JohnnyNOC> i see

[02-Sep-2011 10:33:03] <Sam-I-Am> my power company needs to run zenoss

[02-Sep-2011 10:33:05] <JohnnyNOC> i don't know if i can make that change

[02-Sep-2011 10:33:06] <JohnnyNOC> too hard

[02-Sep-2011 10:33:07] <JohnnyNOC>

[02-Sep-2011 10:33:25] <Sam-I-Am> 3 hour power outage last night... not a light for miles.  called them up "o, we has power outage?"

[02-Sep-2011 10:33:45] <jmp242> lol

[02-Sep-2011 10:33:58] <Sam-I-Am> how can you NOT notice that an entire metro is out

[02-Sep-2011 10:34:12] <Simon4> probably plenty of power in Bangalore

[02-Sep-2011 10:34:25] <Sam-I-Am> haha

[02-Sep-2011 10:34:26] <jmp242> You get a person

[02-Sep-2011 10:34:27] <jmp242> ?

[02-Sep-2011 10:34:32] <jmp242> I get a recording from NYSEG

[02-Sep-2011 10:34:38] <Sam-I-Am> automated call-in, then human callback

[02-Sep-2011 10:34:51] <jmp242> which I love because it for some reason refuses to accept our telephone nuber

[02-Sep-2011 10:35:00] <jmp242> so I have to find the account number somewhere whenever the power is out

[02-Sep-2011 10:35:05] <Sam-I-Am> awesome

[02-Sep-2011 10:35:12] <jmp242> yup

[02-Sep-2011 10:35:43] <jmp242> One of them has only been the same telephone number in the same location since I believe 1948

[02-Sep-2011 10:36:10] <jmp242> So I can see how a land line might confuse them

[02-Sep-2011 10:37:01] <Sam-I-Am> mine knew the cell # i was calling from

[02-Sep-2011 10:37:07] <Sam-I-Am> so i guess their data is correct

[02-Sep-2011 10:37:59] <jmp242> Yea the other one has been a land line in the same location since I think 1979 or os

[02-Sep-2011 10:38:14] <jmp242> land lines confuse NYSEG apparently

[02-Sep-2011 10:38:23] <Sam-I-Am> power companies get confused

[02-Sep-2011 10:40:43] <jmp242> JohnnyNOC: can you expand on the second issue?

[02-Sep-2011 10:40:55] <jmp242> do you get graphs under filesystems for them?

[02-Sep-2011 10:41:14] <konvert> i'm trying with a transform of an event.

[02-Sep-2011 10:41:20] <konvert> this is what i like to do

[02-Sep-2011 10:41:21] <konvert> if " SNMP agent down" in summary.getvalue(): evt_severity=5

[02-Sep-2011 10:41:36] <konvert> is this somewhat correct?

[02-Sep-2011 10:42:32] <jmp242> is that all on one line or two?

[02-Sep-2011 10:42:51] <konvert> on one?

[02-Sep-2011 10:43:15] <konvert> that's the code i've put in the transform field of a mapping

[02-Sep-2011 10:43:28] <jmp242> evt.severity = 5

[02-Sep-2011 10:43:31] <jmp242> not evt_

[02-Sep-2011 10:44:19] <konvert> and then it should work?

[02-Sep-2011 10:46:03] <jmp242> maybe

[02-Sep-2011 10:46:20] <jmp242> I don't know if that check is right

[02-Sep-2011 10:46:34] <konvert> and i don't know python...

[02-Sep-2011 10:46:38] <jmp242> I usually use the formulaton

[02-Sep-2011 10:46:44] <jmp242> (not for summary though)

[02-Sep-2011 10:47:07] <jmp242>  if evt.component == '/mnt/stuff' and evt.severity != 0:

[02-Sep-2011 10:47:08] <jmp242>       evt.severity = 2

[02-Sep-2011 10:48:00] <jmp242> you might want to use

[02-Sep-2011 10:48:52] <jmp242> if " SNMP agent down" in evt.summary

[02-Sep-2011 10:49:03] <jmp242> I think that should work

[02-Sep-2011 10:51:18] <jmp242> JohnnyNOC: Can't repro with a quick remodel of a device I created a mapped network drive on . . .

[02-Sep-2011 10:51:24] <jmp242> seems to work for me

[02-Sep-2011 10:51:32] <konvert> and i have to put this in the transform field of the class snmp?

[02-Sep-2011 10:52:01] <jmp242> You should put it where the events are getting mapped

[02-Sep-2011 10:52:08] <jmp242> event class snmp probably will work

[02-Sep-2011 10:52:14] <jmp242> they are inherited now

[02-Sep-2011 10:53:25] <konvert> and what with the eventclasskey, that is empty in the event

[02-Sep-2011 10:53:58] <konvert> what do i have to put in the eventclasskey field of the mapping?

[02-Sep-2011 10:56:39] <jmp242> heh?

[02-Sep-2011 10:56:42] <jmp242> don't make a mapping

[02-Sep-2011 10:56:47] <jmp242> just put the transform in the event class

[02-Sep-2011 10:58:15] <konvert> where is the transform field in an event class

[02-Sep-2011 10:58:50] <konvert> i can only find an transform field when i edit a mapping

[02-Sep-2011 11:01:13] <jmp242> go to the event class, use the little gear thingy in the lower left to select transform

[02-Sep-2011 11:03:27] <konvert> wow i totaly overlooked it

[02-Sep-2011 11:03:30] <konvert> thx

[02-Sep-2011 11:23:22] <nyeates> We should make the transform UI be on the left-hand menu - i agree its not intuitive to find it in the gear menu

[02-Sep-2011 11:23:53] <nyeates> ...maybe we wanted to hide it from newbies? i dunno

[02-Sep-2011 11:25:28] <dhopp> I think our existing monitoring solution knows it's demise is coming…I plugged in a network cable for one of the zenoss servers in the datacenter and right as the cable clicked into the switch port the light in the DC right above my head flickered

[02-Sep-2011 11:26:11] <JohnnyNOC> jmp242 sorry i was afk

[02-Sep-2011 11:26:52] <nyeates> dhopp: hah. What is in place now?

[02-Sep-2011 11:27:03] <dhopp> nyeates:  it's a home grown application..

[02-Sep-2011 11:28:28] <Hackman238> konvert: evt.severity=5

[02-Sep-2011 11:28:42] <JohnnyNOC> nyeates / dhopp - ever notice a problem with networkDisk being ignored by windows devices?

[02-Sep-2011 11:28:52] <JohnnyNOC> in zFileSystemMapIgnoreTypes

[02-Sep-2011 11:29:09] <dhopp> JohnnyNOC: do you mean like a iscsi disk or something?

[02-Sep-2011 11:29:25] <dhopp> JohnnyNOC: or a disk that is a share on another device?

[02-Sep-2011 11:29:32] <JohnnyNOC> yes

[02-Sep-2011 11:29:36] <JohnnyNOC> like a samba mount

[02-Sep-2011 11:29:44] <JohnnyNOC> or, iscsi is another example sure

[02-Sep-2011 11:29:49] <JohnnyNOC> a filesystem being mapped over the nework

[02-Sep-2011 11:29:51] <JohnnyNOC> network

[02-Sep-2011 11:30:05] <JohnnyNOC> other

[02-Sep-2011 11:30:05] <JohnnyNOC> ram

[02-Sep-2011 11:30:05] <JohnnyNOC> virtualMemory

[02-Sep-2011 11:30:05] <JohnnyNOC> removableDisk

[02-Sep-2011 11:30:05] <JohnnyNOC> floppyDisk

[02-Sep-2011 11:30:06] <JohnnyNOC> compactDisk

[02-Sep-2011 11:30:06] <JohnnyNOC> ramDisk

[02-Sep-2011 11:30:07] <JohnnyNOC> flashMemory

[02-Sep-2011 11:30:08] <JohnnyNOC> networkDisk

[02-Sep-2011 11:30:13] <JohnnyNOC> sorry, didn' expect so many lines

[02-Sep-2011 11:30:19] <JohnnyNOC> but those are my zFileSystemMapIgnoreTypes

[02-Sep-2011 11:31:09] <nyeates> JohnnyNOC: not seen it.... i would search around in http://dev.zenoss.com/trac

[02-Sep-2011 11:31:16] <dhopp> would you want to monitor the disk on the windows server or where it actually resides (in the case of a SMB mount)?

[02-Sep-2011 11:31:47] <JohnnyNOC> where it actually resides

[02-Sep-2011 11:31:51] <JohnnyNOC> i want to only monitor local filesystems

[02-Sep-2011 11:32:02] <dhopp> iscsi should look like a regular disk to snmp (since the underlying networking should be hidden in that case)

[02-Sep-2011 11:32:28] <dhopp> ok…I'm not sure I understand your problem then…you said it's not showing up?  isn't that what you want?

[02-Sep-2011 11:32:35] <dhopp> oh wait..I misread what you asked

[02-Sep-2011 11:32:42] <JohnnyNOC> i have a network drive, mapped to 1000 servers

[02-Sep-2011 11:32:48] <JohnnyNOC> i don't want those 1000 servers to show me that filesystem

[02-Sep-2011 11:32:54] <dhopp> right..I misread what you asked

[02-Sep-2011 11:32:55] <JohnnyNOC> and send alerts, etc.

[02-Sep-2011 11:32:58] <JohnnyNOC> oh, ok

[02-Sep-2011 11:32:59] <dhopp> hrm

[02-Sep-2011 11:35:08] <dhopp> JohnnyNOC: snmp monitoring or wmi?

[02-Sep-2011 11:36:30] <nyeates> i did a search for networkDisk in an internal system....everything is about trying to get network disks to be SEEN :-p

[02-Sep-2011 11:38:45] <nyeates> JohnnyNOC: you should get verbose output of the modeling so run like a manual run of zenmodeller on the device with issues and do it with a -v10

[02-Sep-2011 11:39:27] <nyeates> might show which modeler plugins are being used or not, etc... im not too sure what it will show actually... but it could give a clue

[02-Sep-2011 11:42:16] <JohnnyNOC> dhopp wmi

[02-Sep-2011 11:42:50] <dhopp> hrm…does the wmi modeler use that zProperty?

[02-Sep-2011 11:42:53] <JohnnyNOC> nyeates i think what i'll see is it's not skipping this stuff

[02-Sep-2011 11:43:00] <JohnnyNOC> dhopp i'm not 100% sure

[02-Sep-2011 11:43:05] <JohnnyNOC> i suppose i could have a look

[02-Sep-2011 11:43:17] <dhopp> JohnnyNOC: I'm assuming this Egor's zenpack?

[02-Sep-2011 11:44:08] <JohnnyNOC> dhopp yes

[02-Sep-2011 11:44:14] <JohnnyNOC> using wmi datasource, and wmi perf zenpacks

[02-Sep-2011 11:44:27] <nyeates> JohnnyNOC: click on the monitored drive in the UI. What is shown in the Type field for it?

[02-Sep-2011 11:44:41] <JohnnyNOC> nyeates NTFS

[02-Sep-2011 11:46:18] <nyeates> zenoss might be thinking, for whatever reason, that its a local disk

[02-Sep-2011 11:46:54] <JohnnyNOC> hrm

[02-Sep-2011 11:47:02] <JohnnyNOC> you know, i just RDP'd into the erver and i don't even see that disk right now

[02-Sep-2011 11:47:05] <JohnnyNOC> so i'm going to remodel

[02-Sep-2011 11:47:24] <nyeates> hah yeah....maybe try deleteing the device and readding - if you can afford this

[02-Sep-2011 11:47:58] <JohnnyNOC> dhopp i took a closer look - and the community.wmi.filesystem modeler plugin is using that zProperty

[02-Sep-2011 11:48:35] <JohnnyNOC> which i believe is what matters in this case

[02-Sep-2011 11:50:02] <JohnnyNOC> man

[02-Sep-2011 11:50:08] <JohnnyNOC> i have to delete and re-add? :[  booo

[02-Sep-2011 11:50:14] <nyeates> if remodel doesnt fix it...i contend that the machine is giving incorrect or not-to-spec information to zenoss....either that or zenoss is cacheing or storing the info somewhere it should be...which is why i said to delete and re add

[02-Sep-2011 11:50:33] <nyeates> should be =shouldnt be

[02-Sep-2011 11:50:43] <JohnnyNOC> ok

[02-Sep-2011 11:50:50] <JohnnyNOC> i'm trying to determine how i would go about finding it

[02-Sep-2011 11:50:56] <JohnnyNOC> G: on this device may not be the same as G: on another device

[02-Sep-2011 11:53:56] <JohnnyNOC> well, i think i can figure it out just by looking at some logs

[02-Sep-2011 11:54:12] <JohnnyNOC> but this business of deleting/readding devices to fix kind of sucks

[02-Sep-2011 11:54:31] willwh`afk is now known as willwh

[02-Sep-2011 11:54:37] <JohnnyNOC> i seem to have to do it rather frequently - for example, i see snmp indexes on interfaces seem to change, and RRDs break.

[02-Sep-2011 11:55:11] <JohnnyNOC> and modeling doesn't appear to fix

[02-Sep-2011 12:00:01] [disconnected at Fri Sep  2 12:00:01 2011]

[02-Sep-2011 12:00:03] [connected at Fri Sep  2 12:00:03 2011]

[02-Sep-2011 12:00:19] [zenoss-logger (logger bot) has joined #zenoss]

[02-Sep-2011 12:24:06] <Hackman238> JohnnyNOC: Which devices are exhibiting non persistant ifindexes?

[02-Sep-2011 12:31:48] <photon01> Hi every1

[02-Sep-2011 12:32:13] <photon01> can some1 tell me plz where are mysql tablespaces for history/status tables ?

[02-Sep-2011 12:40:21] <Hackman238> photon01: is mysql on your zenmaster?

[02-Sep-2011 12:40:31] <Hackman238> photon01: If so, su zenoss, mysql -u root -p

[02-Sep-2011 12:40:36] <Hackman238> photon01: use events;

[02-Sep-2011 12:40:42] <Hackman238> photon01: show tables;

[02-Sep-2011 12:40:54] <Hackman238> photon01: status table is current events

[02-Sep-2011 12:41:02] <Hackman238> photon01: history is history

[02-Sep-2011 12:41:12] <Hackman238> photon01: detail contains shared event detail

[02-Sep-2011 12:46:31] <photon01> <Hackman238> : yes, i know. I did a truncate table

[02-Sep-2011 12:46:36] <photon01> <Hackman238> : in order to free space

[02-Sep-2011 12:47:00] <photon01> <Hackman238> : but it seems some tablespace is still present and needs to be removed manually

[02-Sep-2011 12:49:31] <jmp242> photon01: mysql isn't going to free up space when you delete things

[02-Sep-2011 12:49:46] <jmp242> innodb IIRC works this way

[02-Sep-2011 12:49:55] <JohnnyNOC> Hackman238 seemingly random ones

[02-Sep-2011 12:49:58] <jmp242> you actually have to dump the table to a file

[02-Sep-2011 12:50:06] <jmp242> drop the table, and restore your backup

[02-Sep-2011 12:50:09] <jmp242> to reclaim space

[02-Sep-2011 12:50:09] <JohnnyNOC> Hackman238 my assumption is the machine owners are doing something to them which causes the snmp index to change and zenoss just isn't picking up on this

[02-Sep-2011 12:50:19] <JohnnyNOC> and then i get all these alers about 'error reading OIDs...' and broken RRD graphs

[02-Sep-2011 12:50:33] <nyeates> 3.2.0 core coming soon

[02-Sep-2011 12:50:36] <JohnnyNOC> yay

[02-Sep-2011 12:50:41] <JohnnyNOC> why am i afraid of trying to do the upgrade

[02-Sep-2011 12:50:43] <JohnnyNOC>

[02-Sep-2011 12:50:56] <nyeates> heh

[02-Sep-2011 12:50:57] <jmp242> naw, from 3.1 to 3.2 should be easy peasy

[02-Sep-2011 12:51:03] <jmp242> I'm worried about when 4 comes out

[02-Sep-2011 12:51:05] <Hackman238> JohnnyNOC: ifindex persistance issues plauge all monitoring.

[02-Sep-2011 12:51:14] <jmp242> that's been when stuff gets painful in the past lol

[02-Sep-2011 12:51:22] <Hackman238> JohnnyNOC: Chances are they are adding and removing virtual adapters and causing the ifindexes to shuffle

[02-Sep-2011 12:51:23] <JohnnyNOC> jmp242 agreed...  lots of things 'should be'

[02-Sep-2011 12:51:24] <JohnnyNOC>

[02-Sep-2011 12:51:30] <JohnnyNOC> Hackman238 *nod*

[02-Sep-2011 12:51:34] <JohnnyNOC> i just wish a remodel would pick this up

[02-Sep-2011 12:51:36] <JohnnyNOC> ya know?

[02-Sep-2011 12:51:36] <JohnnyNOC>

[02-Sep-2011 12:51:46] <jmp242> no I mean that historically, it's been simple for the point upgrades

[02-Sep-2011 12:51:46] <Hackman238> JohnnyNOC: You need to slap your client and tell them to design their network ahead of time rather than on the fly

[02-Sep-2011 12:51:53] <JohnnyNOC> heh

[02-Sep-2011 12:51:54] <JohnnyNOC> we are the client

[02-Sep-2011 12:51:59] <JohnnyNOC> and we change, a lot

[02-Sep-2011 12:52:06] <photon01> ok, thanks guys

[02-Sep-2011 12:52:11] <JohnnyNOC> sucks for me, but it is what it is

[02-Sep-2011 12:52:14] <Hackman238> JohnnyNOC: Its not possible...theres no real way to match the stuff reliably unless you really trust the interface name

[02-Sep-2011 12:53:06] <Hackman238> JohnnyNOC: I wrote a modeller to handle the problem for Juniper devices having tunnels...the data would get all mized up since the client put up and took down VPN's like everyday

[02-Sep-2011 12:53:17] <Hackman238> *mixed

[02-Sep-2011 12:53:25] <jmp242> mized - new word

[02-Sep-2011 12:53:34] <jmp242> for jupiter devices vpn tunnels lol

[02-Sep-2011 12:53:49] <Hackman238> jmp242: LOL

[02-Sep-2011 12:54:06] <Hackman238> jmp242: Sorry I havent had much input the last day or so- this weekend I'll get right back on the ZCA ops

[02-Sep-2011 12:55:03] <jmp242> Hey, I understand

[02-Sep-2011 12:55:05] <jmp242> work is work

[02-Sep-2011 12:55:34] <jmp242> I don't do anything on weekends lol

[02-Sep-2011 12:58:29] <Hackman238> jmp242: I've just had so many contracts come in

[02-Sep-2011 12:59:14] <Hackman238> jmp242: Some are pretty huge too- I might need to try to delegate and hire some of you as sub-contractors

[02-Sep-2011 12:59:22] <Hackman238> (All Zenoss related)

[02-Sep-2011 13:09:19] <Hackman238> Win on this - http://hackaday.com/2011/09/02/coil-gun-revolver/

[02-Sep-2011 13:17:22] <jmp242> coilgun?

[02-Sep-2011 13:17:30] <jmp242> so it shoots via electricity or something?

[02-Sep-2011 13:22:33] <Hackman238> jmp242: Creates a strong magnetic field to induce a counter field in the load mass to propell the mass forward

[02-Sep-2011 13:23:34] <jmp242> Interesting

[02-Sep-2011 13:25:33] <Hackman238> jmp242: Railgun is differnt. In railgun the electrical load is also the mass load

[02-Sep-2011 13:59:17] <dhopp> how does zenoss determine what order to list the components in?  It's not alphabetical and it doesn't seem to be based on the order of the modeler plugins

[02-Sep-2011 14:15:10] <Hackman238> dhopp: I think its the alpha order of the one to many rel

[02-Sep-2011 14:24:38] willwh is now known as willwh`afk

[02-Sep-2011 14:45:40] <dhopp> Hackman238: you there?

[02-Sep-2011 14:49:45] <Hackman238> dhopp: Hey

[02-Sep-2011 14:49:51] <Hackman238> dhopp: Whats up?

[02-Sep-2011 14:51:34] <dhopp> Hackman238:  hold on..I think i found my problem :-)

[02-Sep-2011 14:53:51] <dhopp> Hackman238:  for a remote collector should the render url be /zport/RenderServer ?

[02-Sep-2011 14:54:30] <Hackman238> dhopp: No, should be http://remotehostname:8091

[02-Sep-2011 14:54:58] <dhopp> do all clients need to be able to resolve that or just the zenoss servers?

[02-Sep-2011 14:56:23] <Hackman238> dhopp: All clients need to be able to hit it

[02-Sep-2011 14:56:36] <Hackman238> dhopp: This method is much faster that rendering off of the zenoss master.

[02-Sep-2011 14:56:47] <Hackman238> dhopp: make sure zenrender is running on the remote collectors

[02-Sep-2011 14:57:46] <dhopp> does it present a problem with multiple collectors?  i.e. if 1 graph is on physical host 1 logical collector 1 and another graph is on physical host 2 logical collector 2?

[02-Sep-2011 15:07:20] <Hackman238> dhopp: You cant create a multigraph report when the devices are on seperate collectors

[02-Sep-2011 15:07:42] <Hackman238> dhopp: There are some patches floating around I think to make it work in v2.5

[02-Sep-2011 15:09:44] willwh`afk is now known as willwh

[02-Sep-2011 15:10:51] <Thinh> is predictive thresholding available for core?

[02-Sep-2011 15:12:18] <Hackman238> Thinh: Yes

[02-Sep-2011 15:12:24] <Hackman238> Thinh: Zepacks page

[02-Sep-2011 15:12:50] <Hackman238> Thinh: You can also write holtwinters graph definitions pretty easily

[02-Sep-2011 15:13:04] <Thinh> on 3.1 right

[02-Sep-2011 15:14:16] <Thinh> can you run zenoss on aVM?

[02-Sep-2011 15:14:20] <Thinh> will it install

[02-Sep-2011 15:16:56] <Hackman238> Thinh: Yes on 3.1

[02-Sep-2011 15:17:04] <Hackman238> Thinh: Zenoss in a VM? Yes and No.

[02-Sep-2011 15:17:15] <Hackman238> Thinh: Yes, but, must be a small instance

[02-Sep-2011 15:17:38] <Hackman238> Thinh: VM = poor disk IO, the most common bottleneck in Zenoss

[02-Sep-2011 15:18:49] <Thinh> just using it for a demo

[02-Sep-2011 15:19:18] <Thinh> is it this one ?

[02-Sep-2011 15:19:18] <Thinh> docs/DOC-10293

[02-Sep-2011 15:20:13] <Hackman238> Thinh: Yep. You must make sure the RRD's are created with the right rrd create commands

[02-Sep-2011 15:21:05] <dhopp> Hackman238:  hmm..that makes me a little sad

[02-Sep-2011 15:21:43] <Hackman238> dhopp: Its not impossible, let me take a look

[02-Sep-2011 15:21:49] <Thinh> ehh how do i install this

[02-Sep-2011 15:21:59] <jmp242> The predictive threshold zenpack?

[02-Sep-2011 15:22:02] <dhopp> Hackman238:  are you saying physical collectors or logical collectors?

[02-Sep-2011 15:22:02] <Thinh> yea

[02-Sep-2011 15:22:10] <jmp242> probably should download it to the Zenoss server

[02-Sep-2011 15:22:12] <jmp242> unzip it

[02-Sep-2011 15:22:27] <jmp242> and then use zenpack --install blah if I recall correctly

[02-Sep-2011 15:23:06] <dhopp> Hackman238:  I will have my physical collectors broken up into 3 logical collectors.  But pretty much a single client's entire production environment will be on the same physical collectors but it may be on different logical collectors

[02-Sep-2011 15:23:23] <Thinh> does it require a restart?

[02-Sep-2011 15:23:47] <Hackman238> dhopp: Ah- you can graph items on the same physical collector even if they are on diff logical collectors

[02-Sep-2011 15:23:59] <dhopp> Hackman238:  ok that makes me a little happier :-)

[02-Sep-2011 15:23:59] <jmp242> well you have to restart zenhub

[02-Sep-2011 15:24:01] <Hackman238> dhopp: You cant graph devices on seperate physical collectors

[02-Sep-2011 15:24:09] <Hackman238> dhopp: Sorry, I should have been clearer

[02-Sep-2011 15:24:13] <jmp242> and zopectl

[02-Sep-2011 15:24:26] <jmp242> did you end up with a .egg?

[02-Sep-2011 15:24:33] <dhopp> Hackman238:  so basically if the render url is different it won't work

[02-Sep-2011 15:24:34] <jmp242> or a tar bundle - cause the installs are different

[02-Sep-2011 15:24:43] <Thinh> tar

[02-Sep-2011 15:24:49] <Hackman238> dhopp: Yep..least not without a work around

[02-Sep-2011 15:25:05] <Thinh> i see it on the gui now

[02-Sep-2011 15:26:28] <dhopp> Hackman238:  I'm trying to figure out why it wouldn't work..I mean zenoss knows which devices are on which collectors

[02-Sep-2011 15:26:35] <Thinh> hack, you said something about creating rrd differently?

[02-Sep-2011 15:26:38] <dhopp> so why not just say "get this data point from here and this one from here"

[02-Sep-2011 15:26:49] <dhopp> it would be slower to render the multigraph report..

[02-Sep-2011 15:27:57] <Hackman238> dhopp: They just never coded it in

[02-Sep-2011 15:28:08] <Hackman238> Thinh: Yeah, it should be in the instructions

[02-Sep-2011 15:28:15] <Hackman238> Thinh: May even be preconfigured

[02-Sep-2011 15:28:25] <Thinh> man i clicked on infrastructure and its taking forever to load

[02-Sep-2011 15:28:25] <Thinh> wtf

[02-Sep-2011 15:28:45] <Bart> hm.. wasn't there going to be an rhel6/centos6 rpm of zenoss 3.2?

[02-Sep-2011 15:29:11] <Hackman238> Bart: Not to my knowledge

[02-Sep-2011 15:29:21] <dhopp> Hackman238:  even in enterprise?  Or is that part of the "enterprise collectors" zenpack that is available for it?

[02-Sep-2011 15:29:29] <Hackman238> Thinh: Is your mysqld hampered?

[02-Sep-2011 15:29:34] <Hackman238> dhopp: Even in ent.

[02-Sep-2011 15:29:42] <dhopp> Hackman238:  that's kind of lame

[02-Sep-2011 15:29:57] <Hackman238> dhopp: No, it has to do with the report system. No one ever put in a ticket for it

[02-Sep-2011 15:30:17] <Hackman238> dhopp: I'd put one in but we dont use that fuctionality, least not yet

[02-Sep-2011 15:30:27] <Hackman238> dhopp: When we do we'll backport the changes to Core

[02-Sep-2011 15:30:33] <Thinh> i only have 3 servers monitoring atm lol

[02-Sep-2011 15:30:59] <Hackman238> Thinh: Yeah, but a misconfigured system thats piling up events can still be slow

[02-Sep-2011 15:31:11] <Hackman238> Thinh: How many datapoints do you collect?

[02-Sep-2011 15:31:24] <Thinh> not alot

[02-Sep-2011 15:31:26] <Thinh> its a new install

[02-Sep-2011 15:31:33] <Thinh> whatever outof thebox + maybe 10 custom

[02-Sep-2011 15:31:37] <Hackman238> Thinh: Is on a VM?

[02-Sep-2011 15:31:42] <Thinh> no its an old box

[02-Sep-2011 15:31:57] <Hackman238> Thinh: Okay. Doesnt have 650 processes on each server to monitor, right?

[02-Sep-2011 15:32:18] <Thinh> is that the magic number

[02-Sep-2011 15:32:39] <Hackman238> Thinh: No, but someone in the forums was asking about a problematic install with 3 devices.

[02-Sep-2011 15:32:54] <Hackman238> Thinh: Had 667 processes per server to monitor

[02-Sep-2011 15:33:10] <Hackman238> Thinh: Causing event floods in zenhub which stops everything from working

[02-Sep-2011 15:33:16] <Thinh> heh

[02-Sep-2011 15:33:38] <Hackman238> Thinh: Anyway, how old is the box?

[02-Sep-2011 15:35:09] <Thinh> dual xeon

[02-Sep-2011 15:35:13] <Thinh> 4gig

[02-Sep-2011 15:35:25] <Thinh> load is really low

[02-Sep-2011 15:35:27] <Thinh> .15

[02-Sep-2011 15:35:30] <Thinh> plenty of disk io

[02-Sep-2011 15:35:40] <Thinh> i restarted mysql and zenoss

[02-Sep-2011 15:35:44] <Thinh> still not loading

[02-Sep-2011 15:35:46] <Thinh> wf

[02-Sep-2011 15:36:43] <dhopp> Thinh: can you stop zenoss and then do a ps aux | grep zenoss before starting it back up?

[02-Sep-2011 15:37:00] <dhopp> wonder if some processes might not be stopping properly

[02-Sep-2011 15:37:04] <Hackman238> Thinh: Sounds like mysql needs to be tweaked for caching and increase the zenhubs -> workers 2, cachsize 10000, pcache 2000, turn on cacheconfig true in all daemons

[02-Sep-2011 15:44:06] <Thinh> zeoctl didnt start

[02-Sep-2011 15:44:11] <Thinh> might be that zenpack =0

[02-Sep-2011 15:46:03] <Hackman238> Thinh: stop zenoss

[02-Sep-2011 15:46:07] <Hackman238> Thinh: Then su zenoss

[02-Sep-2011 15:46:12] <Hackman238> Thinh: zeoctl

[02-Sep-2011 15:46:19] <Hackman238> Thinh: at the prmopt type fg

[02-Sep-2011 15:46:25] <Hackman238> Thinh: What happens?

[02-Sep-2011 15:48:59] <LiamMac> Anyone know how to model an existing device via the JSON api?

[02-Sep-2011 15:55:09] <nyeates> I see someone made the topic change for me!

[02-Sep-2011 15:55:22] <Hackman238> nyeates: Yep, got it covered

[02-Sep-2011 15:55:40] <Hackman238> LiamMac: Not off the top of my head, sorry

[02-Sep-2011 15:57:59] <LiamMac> @hackman238 any suggestions for where to look. It seems to be the missing piece for a project I'm working on

[02-Sep-2011 16:04:44] <Hackman238> LiamMac: Let me take a quick peek

[02-Sep-2011 16:06:23] <Hackman238> LiamMac: REST: ZenModel/Device.py:    def collectDevice(self, setlog=True, REQUEST=None, generateEvents=False)

[02-Sep-2011 16:06:41] <Hackman238> LiamMac: XMLRPC: DataCollector/zenmodeler.py:    def collectDevice(self, device)

[02-Sep-2011 16:19:36] <mattray> enjoying the Zenoss 3.2 release notes

[02-Sep-2011 16:19:42] <mattray> "Support for receipt of SNMPv3 traps"

[02-Sep-2011 16:20:15] <mattray> looking forward to using it on RHEL4

[02-Sep-2011 16:20:50] <Hackman238> mattray: RHEL$?

[02-Sep-2011 16:20:56] <Hackman238> *RHEL4?

[02-Sep-2011 16:21:03] <mattray> it's in the release notes

[02-Sep-2011 16:21:08] <mattray> I suspect they're still testing it

[02-Sep-2011 16:21:12] <LiamMac> @Hackman238 Thanks!

[02-Sep-2011 16:21:28] <Hackman238> mattray: Oh LOL

[02-Sep-2011 16:21:31] <Hackman238> LiamMac: NP

[02-Sep-2011 16:21:59] <Hackman238> LiamMac: When in doubt, cd $ZENHOME/Products, "grep 'def get{whatever}' * -R"

[02-Sep-2011 16:22:02] <Hackman238>

[02-Sep-2011 16:24:43] <mattray> Hackman238: someday I'll get to this: http://tickets.opscode.com/browse/COOK-736

[02-Sep-2011 16:34:25] <Hackman238> mattray: Much appreciated

[02-Sep-2011 16:58:26] <dpetzel> dumb question: when looking at the release notes for 3.2 the defect numbers listed, where can I look for more details on them? It tried dev.zenoss.com but the couple I was interested in dont seem to show up in a search

[02-Sep-2011 16:58:49] <Thinh> hack

[02-Sep-2011 16:58:51] <Thinh>     return FileStorage(config.path, **options)

[02-Sep-2011 16:58:51] <Thinh>   File "/opt/zenoss/lib/python/ZODB/FileStorage/FileStorage.py", line 178, in __init__

[02-Sep-2011 16:58:51] <Thinh>     ltid=ltid, start=start, read_only=read_only,

[02-Sep-2011 16:59:51] <nyeates> mattray: did i put something wrong in the release notes?

[02-Sep-2011 17:00:22] <mattray> not really, just wondering if you're still testing all those old OSs listed in the notes

[02-Sep-2011 17:00:35] <mattray> RHEL4

[02-Sep-2011 17:00:41] <mattray> Ubuntu 8.04

[02-Sep-2011 17:00:57] <nyeates> ohh oh I see. rhel4 - I think we are heh

[02-Sep-2011 17:01:03] <mattray> yikes

[02-Sep-2011 17:01:10] <mattray> and RHEL6?

[02-Sep-2011 17:01:16] <nyeates> :-D

[02-Sep-2011 17:01:34] <nyeates> I beleive its on the blocks for 4.x release of core

[02-Sep-2011 17:01:43] <Thinh> what can i do to troubleshoot zeoctl not starting up

[02-Sep-2011 17:02:21] <mattray> nyeates: nice to see another release of Core, pass on my congratulations to Glenn, QA and the Devs

[02-Sep-2011 17:03:52] <nyeates> Thinh: shutdown everything, watch event.log, startup just zeoctl

[02-Sep-2011 17:04:04] <Thinh> just event log?

[02-Sep-2011 17:05:20] <nyeates> mattray: will do! Glenn says they are going to drop rhel4 and pickup rhel6 for our next major enterprise release

[02-Sep-2011 17:05:26] <Thinh> 2011-09-02T14:10:46 WARNING ZEO.zrpc (508) CW: error connecting to ('localhost', 8100): ECONNREFUSED

[02-Sep-2011 17:05:50] <mattray> nyeates: yay

[02-Sep-2011 17:07:27] <nyeates> mattray: we are for sure moving to github for dev (core and enterprise), and core will be getting ZEP (the overhauled event system) and IPv6 code

[02-Sep-2011 17:07:38] <nyeates> so another few yays

[02-Sep-2011 17:07:58] <mattray> yay++

[02-Sep-2011 17:08:11] <Thinh> welp tht didnt tell me much heh

[02-Sep-2011 17:08:31] <nyeates> Thinh, let me look and see if others have had that issue

[02-Sep-2011 17:08:42] <nyeates> its familiar

[02-Sep-2011 17:10:20] <nyeates> 8100 is port hub connects to zodb database

[02-Sep-2011 17:10:32] <nyeates> check if Data.fs file is around and not messed up

[02-Sep-2011 17:10:42] <nyeates> that is a shot in the dark tho

[02-Sep-2011 17:11:04] <nyeates> mattray: do you have something that pings you when someone says yer name in IRC?

[02-Sep-2011 17:11:45] <mattray> nyeates: indeed I do

[02-Sep-2011 17:12:10] <nyeates> and that is....

[02-Sep-2011 17:13:17] <mattray> oh, I use Adium for my irc client, it has a keyword notification system

[02-Sep-2011 17:13:20] <mattray> 'mattray'

[02-Sep-2011 17:13:21] <mattray> 'matt'

[02-Sep-2011 17:13:23] <mattray> 'mray'

[02-Sep-2011 17:13:42] <mattray> it also does gtalk, yahoo and hipchat

[02-Sep-2011 17:14:04] <Thinh> there is a lock file

[02-Sep-2011 17:14:04] <nyeates> Thinh: i have a possible reason

[02-Sep-2011 17:14:09] <nyeates> Thinh: start mysql

[02-Sep-2011 17:14:26] <Thinh> its running

[02-Sep-2011 17:15:07] <nyeates> make sure its running properly....maybe stop mysql, watch its logs and start it

[02-Sep-2011 17:15:16] <nyeates> another person with this same error, that was their issue

[02-Sep-2011 17:16:13] <nyeates> could be a lock file too tho....there is like a ZEO emergency user related to that i think, if u look it up

[02-Sep-2011 17:16:43] <nyeates> gotta go tho, goodluck

[02-Sep-2011 17:19:50] <Thinh> notihng with mysql

[02-Sep-2011 17:21:05] kocolosk is now known as kocolosk|away

[02-Sep-2011 17:23:26] <Thinh> welp i did a zenrestore with my backup

[02-Sep-2011 17:23:28] <Thinh> its ok now

[02-Sep-2011 17:23:33] <Thinh> i dont know wtf was wrong

[02-Sep-2011 17:23:56] <Thinh> the last thing i did was install that predictive threshold zenpack =0

[02-Sep-2011 17:27:35] willwh is now known as willwh`afk

[02-Sep-2011 17:39:45] willwh`afk is now known as willwh

[02-Sep-2011 19:47:09] kocolosk|away is now known as kocolosk

[02-Sep-2011 20:22:27] willwh is now known as willwh`afk

[02-Sep-2011 20:27:02] willwh`afk is now known as willwh

[02-Sep-2011 20:29:16] kocolosk is now known as kocolosk|away

[02-Sep-2011 20:45:52] willwh is now known as willwh`afk

[03-Sep-2011 00:00:01] [disconnected at Sat Sep  3 00:00:01 2011]

[03-Sep-2011 00:00:02] [connected at Sat Sep  3 00:00:02 2011]

[03-Sep-2011 00:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[03-Sep-2011 00:06:50] willwh`afk is now known as willwh

[03-Sep-2011 01:53:26] willwh is now known as willwh`afk

[03-Sep-2011 12:00:01] [disconnected at Sat Sep  3 12:00:01 2011]

[03-Sep-2011 12:00:01] [connected at Sat Sep  3 12:00:01 2011]

[03-Sep-2011 12:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[03-Sep-2011 13:44:49] willwh`afk is now known as willwh

[03-Sep-2011 14:33:12] <karstensrage> hi

[03-Sep-2011 23:00:56] willwh is now known as willwh`afk

[04-Sep-2011 00:00:01] [disconnected at Sun Sep  4 00:00:01 2011]

[04-Sep-2011 00:00:02] [connected at Sun Sep  4 00:00:02 2011]

[04-Sep-2011 00:00:18] [zenoss-logger (logger bot) has joined #zenoss]

[04-Sep-2011 00:50:37] willwh`afk is now known as willwh

[04-Sep-2011 01:35:53] willwh is now known as willwh`afk

[04-Sep-2011 01:36:18] willwh`afk is now known as willwh

[04-Sep-2011 12:00:02] [disconnected at Sun Sep  4 12:00:02 2011]

[04-Sep-2011 12:00:03] [connected at Sun Sep  4 12:00:03 2011]

[04-Sep-2011 12:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[04-Sep-2011 16:08:37] <rossomaha> hello everyone i need help to install zenoss on linux mint can you help me please?

[04-Sep-2011 16:09:17] <Thinh> what you need

[04-Sep-2011 16:11:46] <rossomaha> i don`t know which repository i need to install in my system. I find in inet binary file but i dont know how to install it

[04-Sep-2011 16:15:09] <rossomaha> I am Russian. Sorry for errors ))

[04-Sep-2011 16:16:42] <rossomaha> Can you help me please?

[04-Sep-2011 16:19:59] <rossomaha> I know i need apache + php+ mysql  + zenoss core. Anybody is alive?

[04-Sep-2011 17:04:03] <willwh> hi rossomaha I am kinda

[04-Sep-2011 17:04:05] <willwh> what's up

[04-Sep-2011 17:04:26] <willwh> rossomaha: you *should* I think, be able to use the debian/ubuntu stack installer

[04-Sep-2011 17:04:32] <willwh> mint is based on debian

[04-Sep-2011 17:05:33] <willwh> http://www.installationwiki.org/Installing_Zenoss#Zenoss_Dependencies

[04-Sep-2011 17:41:08] <rossomaha> by to all

[04-Sep-2011 20:41:41] <Tabmow> How do I set zenoss to by default monitor with v2c instead of v1?

[04-Sep-2011 20:46:50] <Thinh> properties

[04-Sep-2011 20:51:35] <Tabmow> Thinh: I mean when I add new devices.. it sets it to v1 now and I have to change it after the fact which is annoying

[04-Sep-2011 20:51:55] <Thinh> eh

[04-Sep-2011 20:52:16] <Thinh> if you set the properties in a higher lvl it should auto

[04-Sep-2011 20:52:24] <Thinh> linux/server put it under linux

[04-Sep-2011 20:52:25] <Thinh> or whatever

[04-Sep-2011 20:52:40] <Thinh> once oyu drop the server in linux/server it should use the v2c

[04-Sep-2011 21:00:51] <Tabmow> Thinh: ahh ok got ya... I was setting it on the wrong level

[05-Sep-2011 00:00:01] [disconnected at Mon Sep  5 00:00:01 2011]

[05-Sep-2011 00:00:01] [connected at Mon Sep  5 00:00:01 2011]

[05-Sep-2011 00:00:18] [zenoss-logger (logger bot) has joined #zenoss]

[05-Sep-2011 02:50:29] <kipni> goodmorning

[05-Sep-2011 03:00:31] <kipni> did someone already upgrade to 3.2

[05-Sep-2011 03:53:55] <kipni> hello

[05-Sep-2011 04:23:09] <fragfutter> kipni: wrong timezone it gets more lively in the evening

[05-Sep-2011 04:30:41] <kipni> i knew there was something wrong

[05-Sep-2011 04:30:54] <kipni> fragfutter did you upgrade to 3.2 already?

[05-Sep-2011 04:38:04] <danuvius> good morning!

[05-Sep-2011 04:39:29] <kipni> good morning

[05-Sep-2011 04:39:47] <kipni> danuvius did you upgrade to 3.2 yet?

[05-Sep-2011 04:40:03] <danuvius> yes

[05-Sep-2011 04:40:16] <danuvius> at least one of my machines...

[05-Sep-2011 04:40:21] <kipni> can you tell me what the procedure is ?

[05-Sep-2011 04:40:28] <danuvius> I wanted to test out the snmpv3 traps...

[05-Sep-2011 04:42:11] <kipni> can you tell me what the procedure is ?

[05-Sep-2011 04:42:24] <kipni> and do snmv3 traps send more info ?

[05-Sep-2011 04:43:14] <danuvius> Ifollowed the instructions in the installation manual

[05-Sep-2011 04:43:27] <danuvius> community/documentation/official_documentation/installation-guide

[05-Sep-2011 04:43:49] <danuvius> I'm running a debian -stack install and u;dateng ther is quite straight forward ...

[05-Sep-2011 04:44:12] <danuvius> pff i need more morning coffee (making to many typo's)

[05-Sep-2011 04:44:18] <kipni> lol

[05-Sep-2011 04:45:26] <kipni> one more question, how can i know what installation i have?

[05-Sep-2011 04:48:57] <danuvius> kipni: ask the guy who installed it. Else I wouldn't know how to...

[05-Sep-2011 04:49:29] <danuvius> kipni: but their are probably people around here who know though...I'm just not one of them ;-)

[05-Sep-2011 04:50:27] <kipni>

[05-Sep-2011 04:51:07] <Simon4> kipni: what distro are you running it on?

[05-Sep-2011 04:51:19] <kipni> linux

[05-Sep-2011 04:51:38] <kipni> suselinux

[05-Sep-2011 04:52:06] <Simon4> are there files in /usr/local/zenoss/zenoss ?

[05-Sep-2011 04:52:15] <kipni> yes

[05-Sep-2011 04:52:42] <Simon4> that'll be the stack installer then

[05-Sep-2011 04:59:31] <kipni> thx

[05-Sep-2011 05:03:10] <kipni> ok and then next question bin stack installer or debian stack installer

[05-Sep-2011 05:03:28] <Simon4> being suse I would very much doubt it was the debian stack installer

[05-Sep-2011 05:21:32] <kipni> guys in the upgrade document, i have to clean up the cache with two commands, none of the commands work

[05-Sep-2011 05:22:14] <kipni> su zenoss

[05-Sep-2011 05:37:29] <Bart> what would be the best way to customize the zenoss CSS? I tried customizing zenoss.css but apparently that stylesheet isn't loaded on all pages

[05-Sep-2011 05:53:18] <kipni> just did an succesfull upgrade to 3.2

[05-Sep-2011 07:02:46] <hmp> everything working as expected?

[05-Sep-2011 07:18:48] <kipni> yes

[05-Sep-2011 07:27:40] <Tabmow> Is there anyway to delete an OSProcess from a monitoring template?

[05-Sep-2011 07:56:31] fOrsberg is now known as forsberg

[05-Sep-2011 07:58:19] forsberg is now known as fOrsberg

[05-Sep-2011 08:14:19] <Bart> finally managed to create an RPM for centos6, wasn't as easy as I had hoped

[05-Sep-2011 12:00:01] [disconnected at Mon Sep  5 12:00:01 2011]

[05-Sep-2011 12:00:02] [connected at Mon Sep  5 12:00:02 2011]

[05-Sep-2011 12:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[05-Sep-2011 12:05:55] <blerg> If i'd want to somehow monitor the flow of UDP/TCP traffic going through a device, in say the event of a DDoS. Is there any good solution for this? Logs or similar

[05-Sep-2011 12:06:05] <blerg> or how would "you" solve it?

[05-Sep-2011 12:08:21] <hmp> netflow?

[05-Sep-2011 12:34:53] <photon01> hi every1

[05-Sep-2011 12:35:35] <photon01> I need some help with zenoss' mysql database

[05-Sep-2011 16:59:24] * davetoo needs to install/configure Pidgin plugins on this new laptop, in particular to ignore joins/parts

[05-Sep-2011 20:47:14] <Tabmow> Is there anyway to delete an OSProcess from a monitoring template? I deleted it manually for a few devices but 1 of them is still reporting on it for some reason?

[05-Sep-2011 20:48:31] <fus10nx> is there a way to globally set zMaxOIDPerRequest

[05-Sep-2011 20:48:55] <fus10nx> instead of just by device?

[05-Sep-2011 20:50:34] <hmp> in template?

[05-Sep-2011 20:50:54] <Tabmow> fus10nx: set it at the top level

[05-Sep-2011 21:27:18] <fus10nx> what do u mean top level ?

[06-Sep-2011 00:00:01] [disconnected at Tue Sep  6 00:00:01 2011]

[06-Sep-2011 00:00:02] [connected at Tue Sep  6 00:00:02 2011]

[06-Sep-2011 00:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[06-Sep-2011 02:22:24] <fus10nx> anyone around ?

[06-Sep-2011 02:34:08] <konvert> yeah

[06-Sep-2011 02:36:08] <fus10nx> morning

[06-Sep-2011 02:37:52] <fus10nx> where can I set an event trigger ?

[06-Sep-2011 02:38:21] <konvert> morning, what do u mean with an event trigger?

[06-Sep-2011 02:39:12] <konvert> events are automatically created when zenoss get's traps or when a treshold is been exceeded

[06-Sep-2011 02:39:45] <fus10nx> i have a piece of software running on a linux server and its crashed twice in a year...im trying to figure out how I can have it automatically restart

[06-Sep-2011 02:39:55] <fus10nx> if it ever stops or crashes

[06-Sep-2011 02:42:10] <fus10nx> if that's even possible

[06-Sep-2011 02:43:02] <konvert> that's out of the scope of zenoss

[06-Sep-2011 02:43:18] <fus10nx> ah, ok, good to know

[06-Sep-2011 02:43:28] <fus10nx> ill have to try and figure out another solution

[06-Sep-2011 02:44:19] <fus10nx> do you know linux well ?

[06-Sep-2011 02:57:23] <fragfutter> http://libslack.org/daemon/

[06-Sep-2011 03:24:30] <konvert> someone here who can help me drop an event?

[06-Sep-2011 03:37:00] <hmp> zenoss still doesnt support ipv6?

[06-Sep-2011 04:05:01] <konvert> don't know

[06-Sep-2011 04:08:47] <hmp> looks like its cooking for 4.0, which is due to release in december

[06-Sep-2011 04:14:28] <konvert> indeed, 4.0 releases in december

[06-Sep-2011 05:58:48] <konvert> someone who can help me with monitoring my sql server

[06-Sep-2011 07:24:01] <dhopp> hola

[06-Sep-2011 07:24:54] <hmp> cheers

[06-Sep-2011 07:38:23] <konvert> someone who can help me with monitoring my sql server

[06-Sep-2011 08:04:58] <hmp> konvert: mysql monitoring is well documented, do you have any specific questions?

[06-Sep-2011 08:05:59] <konvert> it is not mysql it is mssql

[06-Sep-2011 08:06:49] <konvert> and yes i have some specific questions,

[06-Sep-2011 08:07:06] <konvert> i've succesfully installed all zenpacks

[06-Sep-2011 08:07:23] <konvert> but now i need to create the odbc driver

[06-Sep-2011 08:08:43] <hmp> i dont monitor win machines, sorry

[06-Sep-2011 08:34:50] <Parabola> heyo!

[06-Sep-2011 08:34:52] <Parabola> who missed me

[06-Sep-2011 08:34:57] <konvert> I !

[06-Sep-2011 08:35:10] <Parabola> congrats on the ops rmatte

[06-Sep-2011 08:35:13] <Parabola> did they finally hire you!?

[06-Sep-2011 08:35:18] <konvert> you know something about sql and monitoring ?

[06-Sep-2011 08:35:35] <Parabola> monitoring sql? as in mssql?

[06-Sep-2011 08:35:40] <Parabola>

[06-Sep-2011 08:35:42] <konvert> correct

[06-Sep-2011 08:35:54] <Parabola> i knew it, only microsoft guys call it 'sql'

[06-Sep-2011 08:36:03] <hmp>

[06-Sep-2011 08:36:06] <Parabola> I don't but i might be of assistance, what issues are you having

[06-Sep-2011 08:36:51] <Parabola> and i see a few zenpacks for it :community/zenpacks

[06-Sep-2011 08:37:18] <konvert> i've succesfully installed all zenpacks

[06-Sep-2011 08:37:21] <konvert> but now i need to create the odbc driver

[06-Sep-2011 08:37:31] <konvert> freetsd or something

[06-Sep-2011 08:37:42] <konvert> and there my knoledge ends

[06-Sep-2011 08:40:04] <Gat0rvean> Is it possible to create event transforms with very little knowledge of Python?

[06-Sep-2011 08:40:34] <Parabola> konvert: link to where you read this

[06-Sep-2011 08:41:18] <Gat0rvean> I basically have traps from several devices that are getting mapped into /Unknown, and I can see in the Details pane the traps contain the exact error I need (t1 loss of signal) as well as the port, but Zenoss is not telling me any of that

[06-Sep-2011 08:41:38] <Gat0rvean> or it's not shooting out error emails

[06-Sep-2011 08:57:56] <Sam-I-Am> Gat0rvean: transforms are pretty easy without a lot of python knowledge

[06-Sep-2011 08:59:22] <konvert> Parabola docs/DOC-3391

[06-Sep-2011 09:00:57] <Gat0rvean> Sam-I-Am: I'm having trouble figuring out what the components I need zenoss to grab and report on are

[06-Sep-2011 09:00:59] <jmp242> Gat0rvean: if they're ending up in /Unknown

[06-Sep-2011 09:01:13] <jmp242> you should probably create an event mapping through the web GUI

[06-Sep-2011 09:01:18] <jmp242> before messing with a transform

[06-Sep-2011 09:02:15] <Gat0rvean> jmp242:  My first step was mapping them to /Net/Link and setting the trap that is "Alarm Clear" to clear the "Alarm Set" trap, my issue now, is it's just telling me "ThisDevice" has "this Alarm" I need it also to tell me what port now

[06-Sep-2011 09:02:43] <Gat0rvean> so it will do "This Device" has "this alarm" and then "this device" is "now clear", but it won't grab the port from the trap

[06-Sep-2011 09:03:40] <konvert> parabola found the document?

[06-Sep-2011 09:07:51] <konvert> Gat0rvean do u have the port number in your trap ?

[06-Sep-2011 09:08:11] <Gat0rvean> konvert: I do

[06-Sep-2011 09:10:16] <Gat0rvean> konvert: however, it's not listed under "component" as I'm used it, it's got it's own section called "adMX2800StatusDSXIndex", which is vendor specific (it's an Adtran MX 2800)

[06-Sep-2011 09:24:50] <hmp> any1 using polling on ifOperStatus?

[06-Sep-2011 09:25:07] <Parabola> konvert: http://www.freetds.org/userguide/install.htm

[06-Sep-2011 09:25:12] <Parabola> looks pretty straight forward

[06-Sep-2011 09:25:59] <dhopp> how does Zenoss model 'localhost'?  In a test system it detects the various components but I haven't configured snmp on it yet...

[06-Sep-2011 09:32:46] <konvert> parabola is this for windows ?

[06-Sep-2011 09:33:30] <Parabola> no

[06-Sep-2011 09:34:08] <Parabola> konvert: you need to set this up on your zenoss server

[06-Sep-2011 09:37:19] <JohnnyNOC> anyones 3.2.0 server blow up?

[06-Sep-2011 09:37:19] <JohnnyNOC> :[

[06-Sep-2011 09:37:22] <JohnnyNOC> er, :]

[06-Sep-2011 09:37:32] <konvert> oh ok totaly misunderstood, so i set this up on my zenoss server, and what do i have to set up on my machine i want to monitor?

[06-Sep-2011 09:37:33] <Parabola> that isnt a technical description of a problem

[06-Sep-2011 09:37:52] <Parabola> konvert:  you need to the odbc connection to connect to the MSSQL server.

[06-Sep-2011 09:37:55] <klone> JohnnyNOC - what color was the smoke?

[06-Sep-2011 09:38:14] <Parabola> you don't setup anything on the machine you want to monitor, you just connect to it.

[06-Sep-2011 09:38:28] <Parabola> konvert: sounds like you might want to google "odbc connector"

[06-Sep-2011 09:38:39] <JohnnyNOC> klone hahaha not mine..  was just asking

[06-Sep-2011 09:38:47] <JohnnyNOC> considering giving the upgrade a go tongiht

[06-Sep-2011 09:39:15] <klone> JohnnyNOC - ah, ok. good to hear.

[06-Sep-2011 09:40:06] <JohnnyNOC> figure part of my due diligence would be to ask in here how peoples experiences with 3.2.0 have been these short few days

[06-Sep-2011 09:42:15] <Simon4> JohnnyNOC: it killed my dog and slept with my wife

[06-Sep-2011 09:42:52] <Parabola> haha

[06-Sep-2011 09:43:26] <JohnnyNOC> Simon4 hahahah

[06-Sep-2011 09:44:47] <konvert> parabola pc

[06-Sep-2011 09:45:39] <Parabola> konvert: i'm sorry mate I don't have time to walk you through that right now

[06-Sep-2011 09:45:48] <Parabola> konvert: its not hard, sounds like you're not really a unix guy

[06-Sep-2011 09:45:51] <konvert> ok thx anyway

[06-Sep-2011 09:45:59] <Parabola> konvert: google "install software from source on linux"

[06-Sep-2011 09:46:03] <Parabola> its pretty darn easy

[06-Sep-2011 09:46:42] <Parabola> theres nothing special about the install for this, get source, extract, (have a compiler like gcc installed) "./configure" "make" "make install"

[06-Sep-2011 09:46:44] <Parabola> within that dir

[06-Sep-2011 09:47:32] <konvert> i'm not good at linux, and my boss makes me doing this

[06-Sep-2011 09:48:32] <danuvius> good afternoon

[06-Sep-2011 09:48:50] <konvert> good afternoon

[06-Sep-2011 09:49:58] <Parabola> konvert yeah again i wish i had time right now to help, i'm stuck in a shallow grave too i'm doing some java stuff atm

[06-Sep-2011 09:50:18] <Parabola> konvert: just ask in here for help, or in your linux distro's support channel, it really isnt bad

[06-Sep-2011 09:50:55] <illsci> morning

[06-Sep-2011 09:50:58] <konvert> ok will try

[06-Sep-2011 09:51:02] <danuvius> I'm having in issue importing my self developed zenpack. I exported it, but now it complaints when I try to import it at another installation...

[06-Sep-2011 09:51:11] <SDuensin> Morning.

[06-Sep-2011 09:53:53] <JohnnyNOC> man

[06-Sep-2011 09:54:11] <JohnnyNOC> have you guys seen the lonely island video with michael bolton?

[06-Sep-2011 09:54:20] <JohnnyNOC> i just finished laughing/crying for like 10 minutes

[06-Sep-2011 09:54:40] <danuvius> i created my own eventclass and my own mappings... Both added to the zenpack...

[06-Sep-2011 09:55:55] <illsci> I'm trying to install zenoss on Ubuntu and the zenoss-pack never finishes the install

[06-Sep-2011 09:56:02] <danuvius> but whin installing the zenpack, zenoss complaints: Unable to find context path /zport/dmd/Events/Net/OwnEventClass/instances

[06-Sep-2011 09:56:05] <dhopp> JohnnyNOC: I did it on a test server and basically zencommand monitoring is broken

[06-Sep-2011 09:56:08] <illsci> It just hangs at Setting up zenoss-stack

[06-Sep-2011 09:56:23] <dhopp> JohnnyNOC: snmp oids that start with a '.' won't update the rrd file

[06-Sep-2011 09:56:28] <illsci> I've seen a few posts on this... Is there a better way to install zenoss on Ubuntu?

[06-Sep-2011 09:57:33] <dhopp> JohnnyNOC: those are two separate issues :-)

[06-Sep-2011 09:58:29] <JohnnyNOC> dhopp reading up in the forums now

[06-Sep-2011 09:59:11] <illsci> No one else is running Zenoss on Ubuntu?

[06-Sep-2011 09:59:38] <JohnnyNOC> i'm using zenoss on ubuntu

[06-Sep-2011 09:59:38] <danuvius> yup me...

[06-Sep-2011 09:59:43] <JohnnyNOC> and i installed from the stack installer without problems

[06-Sep-2011 09:59:44] <JohnnyNOC>

[06-Sep-2011 09:59:54] <illsci> I thought that the mysql stuff was included in the zenoss-pack but I've tried setting up zenoss with the package with and without having mysql setup

[06-Sep-2011 10:00:07] <danuvius> same here (took a while though)

[06-Sep-2011 10:00:09] <illsci> Did you have to do any pre-requ stuff

[06-Sep-2011 10:00:21] <illsci> hmm maybe I'm just not waiting long enough

[06-Sep-2011 10:00:28] <illsci> it's been like 5 minutes

[06-Sep-2011 10:01:14] <illsci> Did you setup mysql prior to the install or did you not have it installed when you did the zenoss-stack install?

[06-Sep-2011 10:01:24] <JohnnyNOC> illsci - the stack includes mysql

[06-Sep-2011 10:01:24] <JohnnyNOC> and python

[06-Sep-2011 10:01:31] <JohnnyNOC> all under $ZENHOME

[06-Sep-2011 10:01:39] <JohnnyNOC>  /usr/local/zenoss/mysql

[06-Sep-2011 10:01:40] <illsci> does it sandbox them so they don't interfere with the os's stuff

[06-Sep-2011 10:01:42] <JohnnyNOC>  /usr/local/zenoss/python

[06-Sep-2011 10:01:45] <illsci> werd

[06-Sep-2011 10:01:47] <illsci> hmm

[06-Sep-2011 10:01:50] <JohnnyNOC> not sure about 'sandbox'

[06-Sep-2011 10:01:55] <JohnnyNOC> but it puts them under /usr/local/zenoss

[06-Sep-2011 10:01:56] <illsci> I wonder wtf it's doing

[06-Sep-2011 10:02:01] <JohnnyNOC> and they should not interfere with your warez downloads

[06-Sep-2011 10:02:04] <JohnnyNOC>

[06-Sep-2011 10:02:06] <JohnnyNOC> jk

[06-Sep-2011 10:02:10] <illsci> I'm installing this on a 32bit ec2 instance

[06-Sep-2011 10:02:32] <JohnnyNOC> oh ok, well, you're on your own buddy

[06-Sep-2011 10:02:33] <JohnnyNOC>

[06-Sep-2011 10:02:52] <konvert> guys can someone help, i need to install gcc package and make package on suselinux

[06-Sep-2011 10:08:06] <JohnnyNOC> dhopp so did you manually fix your issue with the OIDs?

[06-Sep-2011 10:08:20] <JohnnyNOC> ie, manually fix some monitoring templates?

[06-Sep-2011 10:09:34] <dhopp> JohnnyNOC: On my test server the only thing I did was on the Linux Device template I changed the memBuffer and memCached to not have the beginning '.' and that got those updating

[06-Sep-2011 10:10:11] <dhopp> JohnnyNOC: I didn't go through every OID to check if they had a '.' since this is just a test box and this probably should be fixed in code

[06-Sep-2011 10:10:23] <JohnnyNOC> dhopp agreed 1000%

[06-Sep-2011 10:10:32] <JohnnyNOC> is nyeates doing the regression testing here?

[06-Sep-2011 10:10:38] <JohnnyNOC>

[06-Sep-2011 10:10:41] <dhopp> lol

[06-Sep-2011 10:11:00] <JohnnyNOC> hehehe jk jk

[06-Sep-2011 10:11:01] <dhopp> I have no idea how the zencommand problems got through QA though…that is really bad...

[06-Sep-2011 10:11:19] <JohnnyNOC> yea

[06-Sep-2011 10:11:29] <JohnnyNOC> i don't want to speculate, but it is disappointing of course

[06-Sep-2011 10:11:33] <JohnnyNOC> just hope they get some patches out soon

[06-Sep-2011 10:12:57] <dhopp> it's kind of bummer because the UI is more streamlined and has some nice fixes to it..and zenhub is supposed to have some fixes to make it handle large number of events better…but doesn't really matter if monitoring is broken

[06-Sep-2011 10:12:58] <dhopp> heh

[06-Sep-2011 10:15:09] <JohnnyNOC> yea

[06-Sep-2011 10:15:13] * JohnnyNOC slaps nyeates with a large trout

[06-Sep-2011 10:15:21] <dhopp> although I think it's funny that the ANNOUNCEMENT on the front page of community zenoss.org says "Zenoss 3.2.0 is now available" but the ANNOUNCEMENT on the forums says "Zenoss 3.1.0 is now available"..heh

[06-Sep-2011 10:15:22] kocolosk|away is now known as kocolosk

[06-Sep-2011 10:16:01] kocolosk is now known as kocolosk|away

[06-Sep-2011 10:16:41] kocolosk|away is now known as kocolosk

[06-Sep-2011 10:17:41] kocolosk is now known as kocolosk|away

[06-Sep-2011 10:17:58] kocolosk|away is now known as kocolosk

[06-Sep-2011 10:23:49] <illsci> yup

[06-Sep-2011 10:24:01] <illsci> it eventually finished installing... that deb install took like 10 minutes

[06-Sep-2011 10:30:04] <wayne___> illsci: I think it's generating the .pyc files at install time

[06-Sep-2011 10:33:03] <jmp242> Morning all

[06-Sep-2011 10:33:11] <dhopp> jmp242:  morning

[06-Sep-2011 10:34:01] <dhopp> jmp242:  do you know how zenoss models 'localhost'?  I am trying to help out a forum question but it seems 'localhost' gets modeled differently since I didn't even have snmp configured properly and it still modeled 2 CPUs

[06-Sep-2011 10:34:19] <jmp242> no idea - I don't use localhost'

[06-Sep-2011 10:34:26] <konvert> morning jmp242

[06-Sep-2011 10:34:27] <jmp242> as far as I know, I use the FQDN

[06-Sep-2011 10:34:30] <jmp242> for my zenoss server

[06-Sep-2011 10:34:44] <jmp242> I recall using localhost caused issues back in the 2.2ish days

[06-Sep-2011 10:34:48] <dhopp> jmp242:  but when you install zenoss by default it creates a 'localhost' device doesn't it?

[06-Sep-2011 10:34:59] <jmp242> ummm, I think so

[06-Sep-2011 10:35:09] <jmp242> but as I said, I think I changed that a while ago

[06-Sep-2011 10:35:11] <jmp242> let me loo

[06-Sep-2011 10:35:13] <jmp242> look

[06-Sep-2011 10:36:04] <jmp242> are you talking about the collector/

[06-Sep-2011 10:36:05] <jmp242> ?

[06-Sep-2011 10:36:11] <jmp242> or just the local server

[06-Sep-2011 10:36:18] <jmp242> in Infrastructure

[06-Sep-2011 10:36:18] <dhopp> no..the device it creates when you do a fresh install

[06-Sep-2011 10:36:37] <jmp242> oh . . . mmm

[06-Sep-2011 10:36:39] <jmp242> no idea

[06-Sep-2011 10:36:49] <jmp242> I last did a fresh install with 2.0 back 3...4 years ago?

[06-Sep-2011 10:40:21] <konvert> jmp242 i'm trying for 6 hours to monitor my mssql server with MS SQL ODBC Database Monitor

[06-Sep-2011 10:40:37] <konvert> on this page MS SQL ODBC Database Monitor

[06-Sep-2011 10:40:57] <konvert> *docs/DOC-3391#

[06-Sep-2011 10:41:12] <konvert> and it is not working

[06-Sep-2011 10:42:04] <dhopp> konvert: what happens?

[06-Sep-2011 10:42:42] <konvert> i've installed freetds like it should (i think)

[06-Sep-2011 10:43:04] <konvert> i've configured the freetds.conf with my two instances of my sqlhosts

[06-Sep-2011 10:43:15] <konvert> i've installed all necessary zenpacks

[06-Sep-2011 10:43:57] <konvert> i've configured the zMsSqlSrvInstances with the instance in freetds.conf

[06-Sep-2011 10:44:08] <konvert> i've configured zwinuser en zwinpasswd

[06-Sep-2011 10:44:21] <konvert> iand now i try to model the device

[06-Sep-2011 10:44:48] <konvert> 2011-09-06 16:44:34,012 ERROR zen.SQLClient: Received error (('00000', '[ISQL]ERROR: Could not SQLConnect')) from query: sp_helpdb

[06-Sep-2011 10:44:48] <konvert> 2011-09-06 16:44:34,031 ERROR zen.SQLClient: Received error (('00000', '[ISQL]ERROR: Could not SQLConnect')) from query: SELECT CONVERT(Char(128), SERVERPROPERTY('InstanceName')) AS InstanceName,

[06-Sep-2011 10:45:14] <dhopp> konvert:  you don't need every sql server you want to monitor in freetds.conf

[06-Sep-2011 10:45:32] <dhopp> did you install unix odbc?

[06-Sep-2011 10:45:37] <dhopp> or pyodbc?

[06-Sep-2011 10:45:50] <dhopp> (pyodbc has to be installed as the zenoss user other wise it doesn't work)

[06-Sep-2011 10:46:23] <konvert> i've only installed freetds

[06-Sep-2011 10:46:24] <dhopp> what version of zenoss and what OS are you using?

[06-Sep-2011 10:46:40] <konvert> zenoss 3.2 on suselinux

[06-Sep-2011 10:46:59] <jmp242> mmm, wouldn't use 3.2 if I were you ...

[06-Sep-2011 10:47:47] <hmp> why not?

[06-Sep-2011 10:47:55] <dhopp> hmp: because it's broken

[06-Sep-2011 10:48:08] <hmp> how?

[06-Sep-2011 10:48:10] <konvert> dhopp in the doc there is not mentioned about pyodbc

[06-Sep-2011 10:48:43] <dhopp> hmp: it seems that if an snmp oid starts with a '.' zenoss doesn't update the rrd file..and so your graphs don't get updated

[06-Sep-2011 10:49:01] <dhopp> hmp: also for zencommand if you have a datasource that has more then one data point only the first data point gets updated

[06-Sep-2011 10:49:19] <hmp> dhopp: thanks

[06-Sep-2011 10:49:30] <hmp> i almost upgradet my 1200+ device production today

[06-Sep-2011 10:49:34] <hmp> that was close

[06-Sep-2011 10:49:47] <konvert> 6:45] <dhopp> konvert:  you don't need every sql server you want to monitor in freetds.conf why not?

[06-Sep-2011 10:50:06] <dhopp> konvert: you have to have either pyodbc or unix-odbc.  as the zenoss user try doing 'easy_install-2.6 pyodbc' (without the quotes)

[06-Sep-2011 10:50:22] <dhopp> konvert: because you don't….freetds is just used to get the driver

[06-Sep-2011 10:50:45] <dhopp> hmp: I would DEFINITELY do upgrades in a test system

[06-Sep-2011 10:51:07] <dhopp> hmp: how many zenoss servers do you have monitoring 1200 devices?

[06-Sep-2011 10:51:10] <dhopp> 1? 2?

[06-Sep-2011 10:51:16] <konvert> so what to put in that freetds.conf

[06-Sep-2011 10:51:39] <hmp> dhopp: 2

[06-Sep-2011 10:52:03] <hmp> dhopp: i'm trying to get interface polling to work on my test setup

[06-Sep-2011 10:52:25] <hmp> so i moved upgrade until i get that working

[06-Sep-2011 10:52:38] <hmp> somehow, the doc i found on community isnt working

[06-Sep-2011 10:52:41] <konvert> dhopp install pyodbc with zenoss user gives an error

[06-Sep-2011 10:52:54] <dhopp> konvert: what error?

[06-Sep-2011 10:53:22] <konvert> setup script exited with error: command 'gcc' failed with exit status 1

[06-Sep-2011 11:01:51] <konvert> dhopp can help me with the error?

[06-Sep-2011 11:07:41] <wayne___> konvert: any errors before that one?

[06-Sep-2011 11:09:27] <dhopp> konvert: you are probably missing a dependancy..I think you need to install unix-odbc-dev (not sure what it's called on suse but you should be able to get it through yast or whatever suse uses)

[06-Sep-2011 11:11:01] <konvert> unixodbc-devel?

[06-Sep-2011 11:11:55] <dhopp> that's probably it

[06-Sep-2011 11:12:04] <konvert> installed it same error

[06-Sep-2011 11:12:28] <dhopp> what version of suse?

[06-Sep-2011 11:14:52] <konvert> trying to find the version, where should i look?

[06-Sep-2011 11:15:02] <hmp> you should really put your whole build log on something like pastebin

[06-Sep-2011 11:18:18] <konvert> dhopp openSUSE 11.4 "Celadon"

[06-Sep-2011 11:24:53] <konvert> dhopp still there?

[06-Sep-2011 11:25:08] <dhopp> I am..but I have a production issue that I'm trying to work through righ tnow

[06-Sep-2011 11:27:00] <konvert> ok np, i need to go now, (europe here) can we go further on this issue tomorrow?

[06-Sep-2011 11:29:35] <ashp> i wonder if zenoss can get me a test drive of service dynamics within about 4 hours so i can show it to my boss at a meeting about moving from nagios to zenoss this afternoon, hmm

[06-Sep-2011 11:29:37] <davetoo> Don't depend on one person in here   Just come back and ask again.

[06-Sep-2011 11:30:01] <konvert> ok davetoo will try thx

[06-Sep-2011 11:37:23] jb______ is now known as jb

[06-Sep-2011 11:37:39] <jb> chet ever hang out here anymore?

[06-Sep-2011 11:38:53] <luizzmizz> hi there

[06-Sep-2011 11:39:20] <luizzmizz> anyone with experience on interfaces for zenoss 3??

[06-Sep-2011 11:39:28] <luizzmizz> i'm stuck...

[06-Sep-2011 11:40:08] <luizzmizz> i'm doing a printermib zenpacks retrieving info about the toners...

[06-Sep-2011 11:40:20] <luizzmizz> the data model is ok and I model correctly the printers

[06-Sep-2011 11:40:33] <luizzmizz> but the panel is killing me

[06-Sep-2011 11:41:34] <jmp242> jb chet occasionally stops by

[06-Sep-2011 11:41:39] <jmp242> but not regularily

[06-Sep-2011 11:59:10] <fus10nx> Hi All....I'm trying to cut down on the number of false/positive alerts I get so I'm wondering how I can get Zenoss to NOT show some of the events (especially from windows event log) such as 'the master browser has recieved a server announcement from computer xxxxx that believes that it is the master browser'

[06-Sep-2011 11:59:18] <fus10nx> or snmp trip snmp_coldStart

[06-Sep-2011 12:00:01] [disconnected at Tue Sep  6 12:00:01 2011]

[06-Sep-2011 12:00:01] [connected at Tue Sep  6 12:00:01 2011]

[06-Sep-2011 12:00:18] [zenoss-logger (logger bot) has joined #zenoss]

[06-Sep-2011 12:00:22] <illsci> whats an appropriate amount of disk space for a production zenoss install that's going to monitor about 100 boxes

[06-Sep-2011 12:10:35] <dhopp> illsci: you could do that on less then 50 or 60 GB probably..unless you change the default RRD

[06-Sep-2011 12:10:54] <dhopp> and even then you probably could

[06-Sep-2011 12:11:57] <hmp> default rrd is 32kB, you multiply that by an aprox number of data points (depending on box type) and add another few GB for the instalation

[06-Sep-2011 12:13:09] <hmp> 1200 cisco/hp/juniper devices (some of them quite big, 6k5, 7k6 and such) uses around 15GB

[06-Sep-2011 12:28:50] <dhopp> hmp: true…and I guess it depends on how many events are stored in the mysql db

[06-Sep-2011 12:29:04] <dhopp> but 50 or 60 will be plenty have plenty of growth :-)

[06-Sep-2011 12:29:41] <dhopp> and have plenty of growth I mean

[06-Sep-2011 12:44:15] <dhopp> fus10nx:  there is a zproperty for windows Event logs for the minimum value to grab…you could try to increase that (or decrease it, not sure which way means more critical)

[06-Sep-2011 13:25:39] <SDuensin> Anyone ever try to psexec into a machine and then psexec out of it to yet another machine?

[06-Sep-2011 13:27:03] <jmp242> SDuensin: no

[06-Sep-2011 13:27:06] <jmp242> why would you?

[06-Sep-2011 13:33:05] <SDuensin> To get implicit login credentials on the *real* target box.

[06-Sep-2011 13:33:41] <SDuensin> If everything was Windows, I could cmdkey /add and stuff my target's credentials into the cache.  Then psexec to it.  If I use -h, I get a high token on that box...

[06-Sep-2011 13:34:19] <SDuensin> However, using Zenoss on a Linux machine with winexe doesn't allow me to do that.  I have to explicitly log in which gives me a medium token and prevents me from installing things because UAC interrupts me...

[06-Sep-2011 13:34:45] <SDuensin> So I was attempting to winexe to a "bounce box" where I could cmdkey/psexec to the true target.

[06-Sep-2011 13:34:50] <SDuensin> Confused yet?  :-)

[06-Sep-2011 13:35:15] <dhopp> SDuensin:  wow you are actually using UAC?  that's brave just by itself

[06-Sep-2011 13:35:31] <SDuensin> Yea, well, not my network.

[06-Sep-2011 13:35:50] <dhopp> (sorry I'm not a fan of UAC)

[06-Sep-2011 13:36:05] <SDuensin> Me either.  Or Windows.  :-)

[06-Sep-2011 13:36:16] <SDuensin> Don't people remember?  The UAC were the BAD GUYS in DOOM!

[06-Sep-2011 13:36:22] <dhopp> lol

[06-Sep-2011 13:41:43] <SDuensin> So you really turn off UAC in a production environment?

[06-Sep-2011 13:42:11] <dhopp> I don't manage windows :-P

[06-Sep-2011 13:42:17] <SDuensin> Lucky bum.

[06-Sep-2011 13:42:51] <dhopp> but when I did and you would research a problem it seems like the solution always was "turn off or down UAC"…too much software wasn't written to work well with it…might be different now

[06-Sep-2011 13:43:17] <SDuensin> Yea, know what wasn't written for it?  WINDOWS!

[06-Sep-2011 13:43:38] <dhopp> I didn't specify which software

[06-Sep-2011 13:44:55] <dhopp> selinux gets in the way a lot too..but at least with it there are ways to update it's configuration to handle it…it just seems with UAC the fix is to play around with a slider that you really don't know what you are changing

[06-Sep-2011 13:46:57] <SDuensin> Pay no attention to the man behind the curtain!

[06-Sep-2011 13:57:07] <jb> if I make changes to a stock enterprise/core zenpack, will those changes get overwritten/deleted during an upgrade?

[06-Sep-2011 13:57:17] <jb> adding additional datasources and whatnot..

[06-Sep-2011 13:58:21] LarsN_ is now known as LarsN

[06-Sep-2011 13:58:27] <dhopp> jb: if you are modifying a template through the UI your changes should stay…if you are modifying underlying code then it will probably get overwritten.  But it may be better to create a copy of the template and then modify that template

[06-Sep-2011 13:58:38] <jb> thanks

[06-Sep-2011 14:05:04] <JohnnyNOC> ZCA members around?

[06-Sep-2011 14:06:06] <JohnnyNOC> just wanted to make a suggestion about something i think would be appropriate for the ZCA to work on

[06-Sep-2011 14:06:25] <JohnnyNOC> if Zenoss INC doesn't - basically a some published guidelines for zenpack developers, etc. about event mappings

[06-Sep-2011 14:06:32] <JohnnyNOC> that would be sweet

[06-Sep-2011 14:24:03] <Gat0rvean> what's the difference between NewDeviceMap and DeviceMap?

[06-Sep-2011 14:29:20] <klone> Gat0rvean - from the comment in DeviceMap.py: "DeviceMap gathers information (sysDescr, sysContact, sysName,

[06-Sep-2011 14:29:20] <klone> and sysLocation) and adds it to the device.

[06-Sep-2011 14:29:20] <klone> To obtain a better OS and hardware manufacturer mapping, use the

[06-Sep-2011 14:29:20] <klone> NewDeviceMap modeler plugin."

[06-Sep-2011 14:53:36] <JohnnyNOC> anyone here utilize sysDesc, sysContact, sysName, or sysLocation?

[06-Sep-2011 14:54:06] <JohnnyNOC> I was talking with my boss about how nice it would be for us to include our asset tag #s in those fields - but seems like it would be a major pain in the butt to get setup for our 1000+ existing devices

[06-Sep-2011 14:54:10] <ashp> Anyone know if it's possible to integrate or add notes to events in zenoss

[06-Sep-2011 14:54:10] <JohnnyNOC> just curious how other people may be using those fields

[06-Sep-2011 14:54:21] <ashp> the idea being you can fill in info for when an alert occurs so your noc knows what to do

[06-Sep-2011 14:54:46] <JohnnyNOC> ashp i don't believe so

[06-Sep-2011 14:54:48] <tvincent> ashp if you doubleclick on a event in the events view a logging option should pop up on the right side of the screen.

[06-Sep-2011 14:55:09] <tvincent> We usually have our NOC put the ticket number in there.

[06-Sep-2011 14:55:19] <JohnnyNOC> tvincent ah nice

[06-Sep-2011 14:55:21] <LiamMac> Can anyone point me to some real beginner info for using the JSON API? I'm having some real trouble defining a python script to remodel a device and I'm pretty it's because eI don't understand the API layout

[06-Sep-2011 14:55:59] <ashp> tvincent: Is that for an individual event or is there any way to attach notes that would appear for that 'type' of event

[06-Sep-2011 14:56:06] <ashp> i think that's more what they are after here

[06-Sep-2011 14:56:15] <ashp> events for 'process matching /regexp/ has this note attached'

[06-Sep-2011 14:56:16] <tvincent> I am not sure how do it for that "type" of event.

[06-Sep-2011 14:56:31] <tvincent> That would be interesting though, KB links inside the event.

[06-Sep-2011 14:57:29] <ashp> that's what we're after, i think it's going to be a lost cause but it would be really useful

[06-Sep-2011 14:58:09] <tvincent> Yes it would be.

[06-Sep-2011 14:58:24] <JohnnyNOC> ashp i assume it can be done with a transform

[06-Sep-2011 14:58:48] <JohnnyNOC> i'm not familiar with this log tvincent was talking about but my assumption is there's some way programmatically you could automagically add a note based on a transform

[06-Sep-2011 14:59:35] <JohnnyNOC> unfortunately looking at the TALES Event attributes I'm not finding it:  docs/DOC-4821

[06-Sep-2011 14:59:40] <tvincent> Go to event console

[06-Sep-2011 14:59:46] <tvincent> double click on a event

[06-Sep-2011 14:59:51] <tvincent> THe portlet opens

[06-Sep-2011 14:59:56] <tvincent> Bottom of it has LOG

[06-Sep-2011 15:00:09] <ashp> I see the log, yeah, it just applies per event, not per .. trigger, i guess, you'd call it

[06-Sep-2011 15:00:14] <JohnnyNOC> ashp you could rewrite the evt.message or evt.summary

[06-Sep-2011 15:00:17] <ashp> JohnnyNOC: I'll have to take a look at transforms

[06-Sep-2011 15:00:26] <ashp> JohnnyNOC: I've never done anything like that - is there any documentation on osmething like that?

[06-Sep-2011 15:00:42] <JohnnyNOC> ashp yes

[06-Sep-2011 15:00:44] <JohnnyNOC> docs/DOC-2554

[06-Sep-2011 15:00:46] <JohnnyNOC> for starters

[06-Sep-2011 15:01:17] <JohnnyNOC> but it sounds like you want to do this based on the 'type' or 'class' of an event and i think that's where you should start poking around

[06-Sep-2011 15:01:38] <JohnnyNOC> the previous link I sent includes the event attributes I know you can modify with a transform

[06-Sep-2011 15:01:46] <JohnnyNOC> perhaps this log thing is also available but not documented..  i dunno.

[06-Sep-2011 15:01:48] <ashp> thanks, that's perfect

[06-Sep-2011 15:02:02] <ashp> we're going to be looking at picking up enterprise to replace nagios

[06-Sep-2011 15:02:04] <JohnnyNOC> so with those 2 links, you should know enough to be dangerous

[06-Sep-2011 15:02:04] <ashp> hurrah!

[06-Sep-2011 15:02:08] <JohnnyNOC>

[06-Sep-2011 15:02:12] <JohnnyNOC> nice

[06-Sep-2011 15:02:16] <JohnnyNOC> i hope you have deep pockets!

[06-Sep-2011 15:02:19] <JohnnyNOC> or few devices

[06-Sep-2011 15:02:20] <JohnnyNOC> :X

[06-Sep-2011 15:02:59] <JohnnyNOC> we manage to do alright with Zenoss Core which is free..  management always likes free

[06-Sep-2011 15:03:00] <JohnnyNOC>

[06-Sep-2011 15:10:45] <jmp242> back from boring meeting

[06-Sep-2011 15:11:38] <dhopp> jmp242:  you mean "informative" meeting

[06-Sep-2011 15:12:08] <jmp242> lol

[06-Sep-2011 15:13:10] <jmp242> a meeting to watch my bosses boss and another department head talk about "cybersecurity" as defined by the NSF that may or may not actually have much to do with IT lol

[06-Sep-2011 15:13:39] <jmp242> I think they just wanted spectators

[06-Sep-2011 15:29:02] <LarsN> Anyone know if/when Core might get released WITH relstorage?

[06-Sep-2011 15:30:07] <dhopp> LarsN: probably not for a while but from what I have heard the implementation in enterprise right now is pretty buggy

[06-Sep-2011 15:30:44] <dhopp> LarsN: from my understanding there isn't a real benefit to relstorage until you get to A LOT of devices…how many devices are you monitoring?

[06-Sep-2011 15:31:28] <LarsN> 200 dedicated servers, and another 200+ virtual machines.  Plus a few hundred URLs with http_monitor

[06-Sep-2011 15:41:27] <tvincent> ashp: We have Enterprise and suppport. Very happy with it.

[06-Sep-2011 15:46:23] <dhopp> tvincent:  I'd be happy if I had it too :-)…I evaluated the enterprise product..was excited about it and then I showed the quote to the people that could do something with it and they looked at me like I had two heads

[06-Sep-2011 15:46:38] <Sashness> LarsN: did you setup multiple http_monitor against the same device?

[06-Sep-2011 15:46:48] <JohnnyNOC> dhopp that was my experience as well

[06-Sep-2011 15:46:50] <tvincent> dhopp: You get what you pay for.

[06-Sep-2011 15:46:53] <tvincent> IN this case.

[06-Sep-2011 15:47:07] <tvincent> Show them a quote for HPOV or Tivoli and then they can have a heart attack.

[06-Sep-2011 15:47:45] <dhopp> tvincent:  agreed..but when you are going from a "free" solution (we have a custom built monitoring system right now) to what zenoss wants it's a tough pill to swallow

[06-Sep-2011 15:48:54] <tvincent> dhopp: Its a case of where you have to compare how many resources you spend maintaining the custom built solutions vs. ZenOSS.

[06-Sep-2011 15:49:45] <JohnnyNOC> here, maintaining the custom built solutions is cheaper than paying for zenoss

[06-Sep-2011 15:49:53] <LarsN> Sashness: in a few cases yes, with local bindings, and specific port numbers.  In other cases where each URL has a unique IP.

[06-Sep-2011 15:49:58] <tvincent> Then your lucky.

[06-Sep-2011 15:50:00] <dhopp> tvincent:  I agree completely..but sometimes it's not viewed that way by upper management because ZenOSS was going to be a check they had to write…the built solution is an assumed cost

[06-Sep-2011 15:50:10] <JohnnyNOC> my salary/yr < what Zenoss would charge us

[06-Sep-2011 15:50:20] <JohnnyNOC> and i can do more than zenoss

[06-Sep-2011 15:50:21] <dhopp> tvincent:  believe me I wanted enterprise but I don't get to make that decision

[06-Sep-2011 15:50:32] <JohnnyNOC> i'm pretty much in the same boat as dhopp

[06-Sep-2011 15:50:42] <tvincent> I am pretty impressed what ZenOSS does open source.

[06-Sep-2011 15:52:03] <tvincent> Alot of Universities and National labs I have worked with in the past have the resources to do everything custom.

[06-Sep-2011 15:53:03] <tvincent> Yea, its always a tough sell. I remember OpsWare. Killer product but expensive as hell.

[06-Sep-2011 15:53:17] <dhopp> tvincent:  what's funny is this…we have OpsWare

[06-Sep-2011 15:54:22] <tvincent> dhopp: how many nodes?

[06-Sep-2011 15:54:41] <dhopp> but we didn't have anything that could do what OpsWare can do in place..so it was easier to sell the cost…we have something in place that works so saying you need to pay $X/year for another solution is a harder sell..

[06-Sep-2011 15:54:58] <dhopp> tvincent:  not sure how many we are licensed for exactly but we have ~1000 servers

[06-Sep-2011 15:55:28] <tvincent> We use Puppet.

[06-Sep-2011 15:55:44] <dhopp> I looked into it..but it doesn't have the Windows support yet

[06-Sep-2011 15:57:09] <tvincent> CFEngine I hear now supports windows. =)

[06-Sep-2011 15:57:09] <JohnnyNOC> any of you guys use an external monitoring service such as Keynote or Webmetrics (Neustar)

[06-Sep-2011 15:57:24] <tvincent> I have in the past. Now I just use Pingdom.

[06-Sep-2011 15:57:46] <JohnnyNOC> tvincent which did you use?  why did you decide just to use pingdom?  didn't need some of the features the other guy soffered?

[06-Sep-2011 15:57:53] <dhopp> JohnnyNOC: We have Webmetrics and gomez..although I don't deal with the configuration of them

[06-Sep-2011 15:58:27] <JohnnyNOC> dhopp we're considering Webmetrics..  any comments?

[06-Sep-2011 15:58:29] <JohnnyNOC> caveats?

[06-Sep-2011 15:58:35] <dhopp> JohnnyNOC: and we use those because somebody decided to use them

[06-Sep-2011 15:58:46] <JohnnyNOC> *nod*

[06-Sep-2011 15:59:05] <tvincent> Keynote I used years ago.

[06-Sep-2011 15:59:12] <Sashness> larsN: thanks

[06-Sep-2011 15:59:20] <tvincent> Pingdom is free or super cheap.

[06-Sep-2011 15:59:23] <dhopp> JohnnyNOC: I can't talk much because I don't deal directly with it..but we get e-mails when a site goes down and get a weekly/monthly report on uptime and response time.

[06-Sep-2011 15:59:27] <tvincent> Gets what I need done.

[06-Sep-2011 15:59:59] <dhopp> JohnnyNOC: the site down e-mails usually coincide with our own monitoring but occasionally our internal monitoring might not see a problem (maybe because of DNS caching or something similar)

[06-Sep-2011 16:00:13] <JohnnyNOC> *nod*

[06-Sep-2011 16:00:18] <Sashness> JohnnyNOC: I use Keynote. We evaluated Neustar but Keynotes pricing for what we do seemed better

[06-Sep-2011 16:00:25] <LarsN> we've been using Websitepulse as a third party, external monitor.

[06-Sep-2011 16:00:33] <tvincent> Monitoring in general, I find, is only as good as the hooks you provide to monitor.

[06-Sep-2011 16:00:48] <dhopp> tvincent:  agreed

[06-Sep-2011 16:00:53] <LarsN> it's neither free, nor cheap. but someone in a desk higher up the row than me made that decision long ago.

[06-Sep-2011 16:01:31] <JohnnyNOC> hrm..  i'm supposed to evaluate and choose one

[06-Sep-2011 16:01:43] <tvincent> JohnnyNOC: save your budget, pingdom =)

[06-Sep-2011 16:01:49] <JohnnyNOC> we've used gomez but apparently disliked them..  and had to get lawyers involved when we wanted to part ways

[06-Sep-2011 16:01:57] <JohnnyNOC> we use monitis now, which is cheap, but i guess sucks

[06-Sep-2011 16:02:16] <JohnnyNOC> tvincent i think part of the problem with pingdom is there are some features we want they don't provide

[06-Sep-2011 16:02:18] <LarsN> I've used the free monitis...

[06-Sep-2011 16:02:26] <JohnnyNOC> for example, one thing we'd like are post-authentication checks

[06-Sep-2011 16:02:26] <LarsN> and yeah, it pretty much sucks.

[06-Sep-2011 16:02:35] <JohnnyNOC> login to the site, click here, do this, do that, etc.

[06-Sep-2011 16:02:54] <JohnnyNOC> i dont think pingdom is a service like that - guess i should read up

[06-Sep-2011 16:02:59] <LarsN> website pulse does a lot of great things.  but it's $12 a month, per url, at 30 minute polling, without any addons.

[06-Sep-2011 16:03:16] <LarsN> addons include, multi-site monitoring,  (ie they check from DC, Chicago, and SF)

[06-Sep-2011 16:03:50] <dhopp> yeah we use the multisite monitoring of webmetrics…and then depending on the client we will use international monitoring too

[06-Sep-2011 16:05:05] <JohnnyNOC> dhopp and you probably have no clue what you pay rightr?

[06-Sep-2011 16:05:05] <JohnnyNOC>

[06-Sep-2011 16:05:18] <JohnnyNOC> based upon my brief experience with both webmetrics and keynote i'm leaning towards webmetrics

[06-Sep-2011 16:05:23] <dhopp> JohnnyNOC: I do not...

[06-Sep-2011 16:05:24] <JohnnyNOC> but that's without considering costs

[06-Sep-2011 16:06:16] <JohnnyNOC> maybe i should just make that recommendation - pingdom, give me the $ we save

[06-Sep-2011 16:06:17] <JohnnyNOC>

[06-Sep-2011 16:07:12] <JohnnyNOC> dhopp you haven't hooked webmetrics into zenoss have you?

[06-Sep-2011 16:07:39] <JohnnyNOC> if i udnerstand correctly i should be able to monitor a log and create alerts

[06-Sep-2011 16:08:00] <dhopp> JohnnyNOC:  not yet :-P

[06-Sep-2011 16:08:08] <LarsN> My problem with the webmetrics is, in some circumstances, with the zenoss server in the same cage as the servers, it's possible to show up, while things are actually unavailable

[06-Sep-2011 16:08:22] <LarsN> where, say the router upstream from ours goes down.

[06-Sep-2011 16:08:52] <JohnnyNOC> LarsN that sounds like an argument for webmetrics (or keynote)

[06-Sep-2011 16:09:06] <JohnnyNOC> maybe i'm misunderstanding you

[06-Sep-2011 16:09:27] <LarsN> JohnnyNOC: I'm just making an observation.

[06-Sep-2011 16:09:28] <LarsN>

[06-Sep-2011 16:09:35] <JohnnyNOC> no problem

[06-Sep-2011 16:09:49] <JohnnyNOC> definitely appreciate the conversation

[06-Sep-2011 16:10:22] <tvincent> LarsN: that is what I like about pingdom. Its like 100 locations around the world.

[06-Sep-2011 16:10:59] <tvincent> But it sounds like someone needs to write a ZenPack for all three.

[06-Sep-2011 16:11:05] <dhopp> the problem with all of them is you need somebody that understand what they are reporting…we have one client that uses gomez on their own and we get graphs when gomez "reports" a problem and then you start looking at the data and the problem will be some third party tracking pixel that is at the end of the web page and has very little impact to the site (and is completely out of our control)

[06-Sep-2011 16:11:07] <JohnnyNOC> tvincent are you on one of these 39.95/mo pingdom plans?

[06-Sep-2011 16:11:50] <JohnnyNOC> seems like there are only 3 options there

[06-Sep-2011 16:12:15] <JohnnyNOC> dhopp you're absolutely right..  my boss laid out some 'nice to haves' but i dont think we'll be able to justify the cost for these things

[06-Sep-2011 16:12:20] <JohnnyNOC> and will probably end up with something like pingdom

[06-Sep-2011 16:13:07] <JohnnyNOC> and i'm likely the only one to actually 'use' it

[06-Sep-2011 16:14:03] <jmp242> OT - anyone here use VPN services?

[06-Sep-2011 16:14:43] <LarsN> jmp242: we run an OpenVPN server if that's what you're asking.

[06-Sep-2011 16:14:48] <JohnnyNOC> we have cisco asas

[06-Sep-2011 16:15:13] <JohnnyNOC> jmp242 have a specific question?

[06-Sep-2011 16:15:19] <jmp242> oh no - I mean like hidemyass.com

[06-Sep-2011 16:15:34] <jmp242> for wifi security for my android

[06-Sep-2011 16:15:41] <jmp242> when I'm out and about

[06-Sep-2011 16:15:56] <LarsN> on a Zenoss related note.  Is there a guide that explains how to set up device hierarchy?  (ie. Router 1 has servers 100-199 behind it / router 2 has servers 200-299 behind it, router 3 has an entire remote site behind it.)  If router 2 fails...  I'd like to only get an alert that Router2 died, rather than router2, plus 200 devices all dided.

[06-Sep-2011 16:16:15] <jmp242> LarsN: not really

[06-Sep-2011 16:16:21] <jmp242> if the pingtree is right

[06-Sep-2011 16:16:28] <jmp242> zenoss is supposed to do layer3 automatically

[06-Sep-2011 16:16:33] <jmp242> but it barely works sometimes

[06-Sep-2011 16:16:40] <jmp242> due to the issue with async pings...

[06-Sep-2011 16:16:41] <LarsN> jmp242: I haven't set anything up on my new Android device yet.  on my N900 I just had openVPN installed and running.

[06-Sep-2011 16:16:49] <jmp242> yea, my phone won't run OpenVPN

[06-Sep-2011 16:16:56] <jmp242> cause I can't root the Droid3

[06-Sep-2011 16:17:00] <jmp242> so I need PPTP ...

[06-Sep-2011 16:18:54] <LarsN> jmp242: https://market.android.com/details?id=de.akvsoft.droid.ovpninstall&feature=search_result plus https://market.android.com/details?id=de.schaeuffelhut.android.openvpn.installer&feature=search_result

[06-Sep-2011 16:18:59] <LarsN> won't work without rooting eh?

[06-Sep-2011 16:19:49] <tvincent> larsN: you can do a event transfom and supress all the alerts if ZenOSS is unable to ping that core router.

[06-Sep-2011 16:20:08] <tvincent> larsN: I do it by putting all my servers into locations and then supress based on the location of the server.

[06-Sep-2011 16:20:47] <tvincent> JohnnyNoc: Yes I am on a plan around that pricing.

[06-Sep-2011 16:21:10] <LarsN> tvincent: rgr.  That was going to be my solution if Zenoss wasn't smart enough to have included similar functionality by default.

[06-Sep-2011 16:21:28] <LarsN> tvincent: at least locations can easily be set via the command line (unlike systems)

[06-Sep-2011 16:21:54] <tvincent> LarsN: It would be nice if they included that in the GUI at least.

[06-Sep-2011 16:22:15] <tvincent> LarsN: I think you can do it via the REST api or dmd pretty easily, systems that is.

[06-Sep-2011 16:22:24] <LarsN> tvincent: and got Priorities working correctly.

[06-Sep-2011 16:23:45] <LarsN> tvincent: currently the Priorities dropdown in the Alerts setion, provide the values from Severity...

[06-Sep-2011 16:25:13] <LarsN> I've lost the link, but I'm pretty sure the bug I filed, got set to "won't fix"

[06-Sep-2011 16:30:58] <klone> LarsN - this one? http://dev.zenoss.com/trac/ticket/7191

[06-Sep-2011 16:33:45] <klone> if that's the one, the filter field you want is 'Device Priority', which reads the values you're expecting to see

[06-Sep-2011 16:38:19] <fus10nx> Is there a way to globally set zWinEventlogMinSeverity

[06-Sep-2011 16:38:30] <fus10nx> so i dont have to do it for each machine. I want the default to be 1 not 2

[06-Sep-2011 16:40:10] <dhopp> You can set it on the device class

[06-Sep-2011 16:40:42] <dhopp> but if you changed it at the device level you will need to tell Zenoss to delete that local property so it will inherit it

[06-Sep-2011 16:40:52] <fus10nx> hmmmmm

[06-Sep-2011 16:40:55] <fus10nx> is that an easy task?

[06-Sep-2011 16:40:58] <dhopp> yeah

[06-Sep-2011 16:41:05] <fus10nx> I just keep getting all this windows event log stuff that triggers alerts that i dont care about

[06-Sep-2011 16:41:10] <fus10nx> its useless to us to be honest

[06-Sep-2011 16:41:20] <fus10nx> is there a way to just ignore certain messages?

[06-Sep-2011 16:41:43] <dhopp> that's a little trickier..you need to deal with tranforms and event classes

[06-Sep-2011 16:41:50] <dhopp> but it can be done

[06-Sep-2011 16:41:56] <fus10nx> hmmmm

[06-Sep-2011 16:41:59] <dhopp> read up on even transformations

[06-Sep-2011 16:42:01] <fus10nx> ok

[06-Sep-2011 16:42:02] <fus10nx> thanks

[06-Sep-2011 16:42:06] <dhopp> err

[06-Sep-2011 16:42:10] <dhopp> event transformations

[06-Sep-2011 16:42:41] <fus10nx> oh man, this is like programming haha

[06-Sep-2011 16:43:02] <dhopp> yeah..that's why I said it's a little trickier

[06-Sep-2011 16:43:03] <dhopp> heh

[06-Sep-2011 16:43:10] <fus10nx> ya this goes beyond my scope

[06-Sep-2011 16:43:41] <fus10nx> the only thing im trying to resolve is not getting 50 emails that are useless

[06-Sep-2011 16:44:14] <dhopp> you may be able to tell zenoss that a certain event is of a certain event class and then just drop all events of that class

[06-Sep-2011 16:45:11] <dhopp> fus10nx:  what you may want to do is set zWinEventlog to False at the Windows device class

[06-Sep-2011 16:45:24] <dhopp> and then on just one server set it to true until you work out how to filter the messages

[06-Sep-2011 16:45:59] <fus10nx> Well I still want the critical errors

[06-Sep-2011 16:46:03] <fus10nx> so I don't wnat to fully do that

[06-Sep-2011 16:46:49] <dhopp> fus10nx:  ok..I didn't know how far along in your implmentation you were

[06-Sep-2011 16:46:54] <dhopp> implementation*

[06-Sep-2011 16:59:21] <fus10nx> sorry

[06-Sep-2011 16:59:24] <fus10nx> ya

[06-Sep-2011 16:59:26] <fus10nx> not very far

[06-Sep-2011 16:59:39] <fus10nx> its just annoying to get all these emails if they aren't relevant

[06-Sep-2011 16:59:52] <fus10nx> i only want to see what's important...telling me that snmp did a cold start....eh, i dont care ! haha

[06-Sep-2011 17:00:27] <willwh> fus10nx: then change it

[06-Sep-2011 17:00:36] <fus10nx> thats what im trying to figure out

[06-Sep-2011 17:00:41] <fus10nx> trying to figure out how to it

[06-Sep-2011 17:05:26] <dhopp> fus10nx:  what is the event class on those snmp cold start messages?

[06-Sep-2011 17:42:39] kocolosk is now known as kocolosk|away

[06-Sep-2011 18:00:45] <tvincent> so have the zenoss offices been evacuated from the fire yet?

[06-Sep-2011 18:05:44] <dhopp> tvincent: huh?

[06-Sep-2011 18:25:16] kocolosk|away is now known as kocolosk

[06-Sep-2011 18:33:28] <mloven> tvincent:  The wildfires weren't terribly close to the Austin office, so I think they're ok.

[06-Sep-2011 19:32:54] <Thinh> is there a feature in zenoss  where zencollector would switch automatically

[06-Sep-2011 19:34:26] <tvincent> dhopp: I was referring to the Austin offices.

[06-Sep-2011 19:58:17] <kocolosk> tried upgrading to 3.2, seems like ssh metrics aren't being collected anymore

[06-Sep-2011 19:58:20] <kocolosk> is that a known issue?

[06-Sep-2011 19:58:53] <kocolosk> saw some comments in the backlog by dhopp and jmp242 about 3.2 being broken, but i think my backlog doesn't go far enough back to see why :/

[06-Sep-2011 19:59:32] <Thinh> ssh metrics?

[06-Sep-2011 20:00:20] <kocolosk> collection of metrics via zencommand

[06-Sep-2011 20:00:59] <kocolosk> i use snmp for the basics but wrote some nagios plugins for non-standard stuff, and i collect metrics from those using ssh

[06-Sep-2011 20:01:18] <kocolosk> the SNMP metrics are still updating, but the ones collected by zencommand are not

[06-Sep-2011 20:02:35] <Thinh> d you know what command specifically

[06-Sep-2011 20:03:36] <kocolosk> oh, it's a custom plugin ... but that reminds me of something

[06-Sep-2011 20:03:47] <Thinh> sometime its sa permission issue after upgrade

[06-Sep-2011 20:04:01] <kocolosk> i seem to remember zenoss wanted the plugin on the path on the server as well as on the client for some reason

[06-Sep-2011 20:04:29] <Thinh> it run it as zenoss user and i think zenoss can only run in libexec right

[06-Sep-2011 20:04:30] <kocolosk> Thinh: yeah, i'll check ... I did check that the zencommand.log isn't puking out any errors or warning

[06-Sep-2011 20:04:53] <kocolosk> yeah, sounds familiar

[06-Sep-2011 20:09:34] <kocolosk> Thinh: bingo, adding those scripts back into $ZENHOME/../common/libexec did the trick

[06-Sep-2011 20:09:42] <kocolosk> thanks for jogging my memory

[06-Sep-2011 20:20:12] <Thinh> i thikn you can use the PUSH option to push the scripts to zencollector

[06-Sep-2011 20:20:57] <kocolosk> i don't know anything about that

[06-Sep-2011 20:21:54] <kocolosk> also, i spoke too soon -- i have one datapoint from each plugin being collected

[06-Sep-2011 20:22:07] <kocolosk> the others haven't been collected since the 3.2 upgrade

[06-Sep-2011 20:22:18] <Thinh> all your daemons running?

[06-Sep-2011 20:22:27] <kocolosk> afaik

[06-Sep-2011 20:22:38] <kocolosk> did a zenoss-stack restart, all came up cleanly

[06-Sep-2011 20:23:20] <Thinh> try the push changes option

[06-Sep-2011 20:23:24] <Thinh> click on the gadget

[06-Sep-2011 20:23:44] <kocolosk> which gadget?

[06-Sep-2011 20:24:06] <Thinh> drill down to the zencollector host on the GUI, the gadget on the bototm left

[06-Sep-2011 20:24:50] <Thinh> or just scp everything in libexec on your main hub to your zencollector heh

[06-Sep-2011 20:25:02] <kocolosk>

[06-Sep-2011 20:25:27] <dhopp> kocolosk: are these for zencommand data points?

[06-Sep-2011 20:25:29] <kocolosk> my only collector is on localhost

[06-Sep-2011 20:25:32] <kocolosk> dhopp: yup

[06-Sep-2011 20:25:38] <Thinh> oh

[06-Sep-2011 20:25:43] <dhopp> it's a known bug in 3.2

[06-Sep-2011 20:25:46] <kocolosk> getting one datapoint per data source

[06-Sep-2011 20:25:50] <Thinh> thats fucked

[06-Sep-2011 20:25:53] <kocolosk> dhopp: thanks, that's why i pinged you

[06-Sep-2011 20:25:57] <kocolosk> link?

[06-Sep-2011 20:25:58] <dhopp> http://dev.zenoss.org/trac/ticket/7860

[06-Sep-2011 20:26:01] <kocolosk>

[06-Sep-2011 20:26:11] <Thinh> good thing i didnt upgrade heh

[06-Sep-2011 20:26:11] <kocolosk> ty

[06-Sep-2011 20:26:41] <dhopp> Thinh: yeah there are a couple of nasty bugs…Zenoss is working on them and I would assume a patch will be coming fairly quickly

[06-Sep-2011 20:27:07] <kocolosk> "ZenCommand isn't even attempting an update of a command datasource that has multiple datapoints. The first datapoint is updated and then it stops."

[06-Sep-2011 20:27:20] <dhopp> kocolosk:  yup..

[06-Sep-2011 20:27:38] <kocolosk> alrighty

[06-Sep-2011 20:27:49] <kocolosk> glad to hear it's a known issue and P1

[06-Sep-2011 20:27:54] <dhopp> I did a fresh install of 3.2 to verify it wasn't a problem with a 3.1 to 3.2 upgrade

[06-Sep-2011 20:28:00] <kocolosk> ah, thanks

[06-Sep-2011 20:28:20] <dhopp> problem is still there…I briefly talked with one of the devs earlier today...

[06-Sep-2011 20:28:38] <dhopp> don't know of an eta on a fix..but I would imagine fairly quickly

[06-Sep-2011 20:32:19] <kocolosk> i guess i should've worked a bit harder to build a "bugs in 3.2" trac query: http://dev.zenoss.org/trac/query?status=assigned&status=backlog&status=closed&status=fixrejected&status=new&status=verification&order=priority&version=3.2.0

[06-Sep-2011 20:53:00] <Tabmow> If I have some devices which are still showing events for OSProcesses that I have already deleted, how can I get fix that?

[06-Sep-2011 20:55:01] <dhopp> Tabmow:  are the events recent?  meaning is the LastSeen in the last few minutes

[06-Sep-2011 20:55:02] <dhopp> ?

[06-Sep-2011 21:12:29] <Tabmow> Sure is...

[06-Sep-2011 21:12:35] <Tabmow> oops his gone

[07-Sep-2011 00:00:01] [disconnected at Wed Sep  7 00:00:01 2011]

[07-Sep-2011 00:00:01] [connected at Wed Sep  7 00:00:01 2011]

[07-Sep-2011 00:00:01] [disconnected at Wed Sep  7 00:00:01 2011]

[07-Sep-2011 00:00:02] [connected at Wed Sep  7 00:00:02 2011]

[07-Sep-2011 00:00:20] [zenoss-logger (logger bot) has joined #zenoss]

[07-Sep-2011 02:39:32] <konvert> goodmorning

[07-Sep-2011 03:25:21] <konvert> anyone online yet (or still online?)

[07-Sep-2011 05:55:23] Kiarash_MON is now known as whyzgeek

[07-Sep-2011 07:38:27] <konvert> hi dhopp

[07-Sep-2011 08:43:24] klone is now known as psyklone

[07-Sep-2011 08:54:27] <Hackman238> Morning all

[07-Sep-2011 08:55:05] <konvert> morning Hackman238

[07-Sep-2011 08:55:24] <konvert> can i ask you something

[07-Sep-2011 08:55:32] kocolosk is now known as kocolosk|away

[07-Sep-2011 08:57:45] <JohnnyNOC> good morning SWS the 2nd

[07-Sep-2011 08:59:32] <konvert> goodmorning

[07-Sep-2011 09:01:33] <konvert> is it possible to drop an event based on the component?

[07-Sep-2011 09:03:15] <konvert> i have an event that informs me of an cdrom in one of the devices that is full, so the capacity of the cdrom is full, i don't need to know that.

[07-Sep-2011 09:03:56] <konvert> the component is /media/XenServer Tools

[07-Sep-2011 09:04:20] <konvert> maybe i can drop every event from that perf class that has media in the component?

[07-Sep-2011 09:06:21] kocolosk|away is now known as kocolosk

[07-Sep-2011 09:09:24] <konvert> someone with an idea?

[07-Sep-2011 09:10:20] <Simon4> if "media" in evt.component:

[07-Sep-2011 09:10:26] <Simon4>     evt._action = "drop"

[07-Sep-2011 09:10:39] <Simon4> would about do it

[07-Sep-2011 09:10:53] <Simon4> presuming you're looking for the string "media"

[07-Sep-2011 09:11:09] <Simon4> hurl that in the transform space for the appropriate event class

[07-Sep-2011 09:15:33] <konvert> Simon do i need to add some wildcards in the string

[07-Sep-2011 09:16:12] <Simon4> konvert: no, that does a substring search

[07-Sep-2011 09:16:30] <Simon4> so it'll match "removable media" or "media things" etc

[07-Sep-2011 09:25:33] <Gat0rvean> Can anyone explain why the graph in Zenoss claims zenperfsnmp cycle times are avg. 52.81 while the logs show the cycle lasted 260ish?

[07-Sep-2011 09:28:52] <jmp242> Gat0rvean: it's an average - perhaps some of the cycle times are much less?

[07-Sep-2011 09:29:41] <Gat0rvean> jmp242: yeah, I just noticed that actually, on the 25th of August some event happened that made them go WAYY up, but I have no logs that far back fml

[07-Sep-2011 09:32:35] <konvert> i have this error in my eventlog : The threshold Memory in template /zport/dmd/Devices/Server/Windows/4%20Processors/rrdTemplates/Device has caused an exception its from the component zenhub and the device is my zenoss machine

[07-Sep-2011 09:32:41] <konvert> what can be the error?

[07-Sep-2011 09:33:21] <konvert> more info : User-supplied Python expression (here.getRRDValue('MemoryTotal') * 0.9) for maximum value caused error: ['Memory_MemoryUsed']

[07-Sep-2011 09:50:04] <whyzgeek> Gat0rvean I recall last time I check the code, it is always divided by 3

[07-Sep-2011 09:50:28] <whyzgeek> I think the logic can be it make it something like a percentage

[07-Sep-2011 09:50:43] <whyzgeek> so when it shows 100 on the graph

[07-Sep-2011 09:50:55] <whyzgeek> that means you got to 300sec collection cycle

[07-Sep-2011 09:52:28] <whyzgeek> so 260 should show you current as 86

[07-Sep-2011 09:53:03] <SDuensin> Good morning.

[07-Sep-2011 10:09:34] <konvert> someone who can help me?

[07-Sep-2011 10:09:40] <konvert> [15:32] <konvert> i have this error in my eventlog : The threshold Memory in template /zport/dmd/Devices/Server/Windows/4%20Processors/rrdTemplates/Device has caused an exception its from the component zenhub and the device is my zenoss machine

[07-Sep-2011 10:09:40] <konvert> [15:32] <konvert> what can be the error?

[07-Sep-2011 10:09:40] <konvert> [15:33] <konvert> more info : User-supplied Python expression (here.getRRDValue('MemoryTotal') * 0.9) for maximum value caused error: ['Memory_MemoryUsed']

[07-Sep-2011 10:28:33] <konvert> someone else having problems with the performance graphs of windows servers after upgrade to 3.2 ?

[07-Sep-2011 10:36:00] <JohnnyNOC> i haven't upgraded to 3.2 but i read there seemed to be an issue with OIDs that start with a .

[07-Sep-2011 10:36:05] <JohnnyNOC> perhaps that's the issue you're seeing?

[07-Sep-2011 10:36:43] <konvert> i have a red cross for every graph (like an missing image) when going to graphs

[07-Sep-2011 10:38:39] <konvert> maybe a reinstall of the zenpack could help?

[07-Sep-2011 10:51:51] <konvert> anyone online and has time?

[07-Sep-2011 10:53:11] <wayne___> konvert: online yes, time maybe

[07-Sep-2011 10:54:02] <konvert> i've lost my graphs on my windows servers after update to 3.2

[07-Sep-2011 10:54:21] <illsci> Do any of you guys host zenoss on Ec2?

[07-Sep-2011 10:54:53] <illsci> I'm curious about the instance types you use and how you keep data around...  Do you use ebs backed images?  Do you mount EBS at certain places in the file systems?

[07-Sep-2011 10:55:28] <wayne___> konvert: are the graphs gone altogether or are you just not getting any more data points?

[07-Sep-2011 10:58:21] <konvert> i just have the heading with buttons to zoomin and out and then i have a red cross (like a missing image)

[07-Sep-2011 10:58:49] <wayne___> any errors in the logs?

[07-Sep-2011 10:58:58] <konvert> where to see?

[07-Sep-2011 10:59:39] <wayne___> there should be a log directory under the zenoss install directory

[07-Sep-2011 10:59:55] <wayne___> in the case of the .deb stack it's /usr/local/zenoss/zenoss/log/

[07-Sep-2011 11:02:28] <konvert> in what log i have to look? zenwin, zenprocess, zenstatus,...

[07-Sep-2011 11:04:42] <wayne___> I'm not sure.  Start with the one that was written to most recently

[07-Sep-2011 11:07:14] <konvert> WARNING zen.snmpPerfConfig: mib lookup for .1.3.6.1.2.1.2.2.1.8 failed

[07-Sep-2011 11:07:15] <konvert> c

[07-Sep-2011 11:07:20] <konvert> can that be the error?

[07-Sep-2011 11:08:41] <dhopp> konvert: verify your rrds are still there in $ZENHOME/perf/Devices/[whatever your device is]

[07-Sep-2011 11:10:24] <konvert> yes still there

[07-Sep-2011 11:11:02] <rmatte> konvert: the update to 3.2 has several very critical bugs in it

[07-Sep-2011 11:11:13] <konvert> strange, sometimes the graph is there but when i click on it it is gone and then again a red cross

[07-Sep-2011 11:11:15] <rmatte> including graphing issues

[07-Sep-2011 11:11:38] <rmatte> it's strongly recommended not to upgrade until the bugs get addressed

[07-Sep-2011 11:11:45] <rmatte> they were identified a couple of days ago

[07-Sep-2011 11:12:19] <nexex> thank goodness for zenrestore

[07-Sep-2011 11:12:33] <konvert> is a restore easy?

[07-Sep-2011 11:12:49] <dhopp> konvert: since you upgraded have you closed your browser down?  This almost sounds like a browser cache problem

[07-Sep-2011 11:12:50] <rmatte> well, I don't know how well zenrestore is going to work after an upgrade... that's more meant for saving backups of the current version

[07-Sep-2011 11:13:11] <rmatte> try dhopp's solution... maybe clear your cache

[07-Sep-2011 11:13:37] <rmatte> but like I said, there are serious issues with 3.2, for instance, any snmp datasource that has an OID starting with a period won't work

[07-Sep-2011 11:13:44] <rmatte> command datasources don't work either

[07-Sep-2011 11:13:57] <rmatte> the scripts exit with code 3 or something

[07-Sep-2011 11:14:26] <dhopp> rmatte: I actually did some more debugging..they don't exit with code 3…they just don't parse all of the data points..

[07-Sep-2011 11:14:41] <rmatte> well either way, that's a big problem

[07-Sep-2011 11:14:43] <rmatte> lol

[07-Sep-2011 11:14:56] <rmatte> not fit for release

[07-Sep-2011 11:15:00] <dhopp> rmatte:  another community member found that reverting Auto.py to version 3.1 it fixes the command data source problem..however it's unknown what that means to the rest of the upgrade

[07-Sep-2011 11:15:05] <dhopp> rmatte:  agrred

[07-Sep-2011 11:15:08] <dhopp> agreed*

[07-Sep-2011 11:15:14] <rmatte> I'm blown away that that made it through QA, I never had a chance to properly beta test 3.2, been really busy with work

[07-Sep-2011 11:15:42] <rmatte> That seems like some really basic QA that's lacking (if any was performed at all)

[07-Sep-2011 11:16:36] <dhopp> rmatte:  yeah I'm surprised too…the oid starting with a '.' I might give them a pass because that was kind of hard to find..but the command datasource is a big one and very obvious that it didn't work (the fix may not be easy but detecting it is)

[07-Sep-2011 11:16:49] <rmatte> I'm going to bring this up at the next ZCA meeting... we should almost request that no core releases hit public prior to testing and approval by ZCA members

[07-Sep-2011 11:17:03] <konvert> do u guys have a doc to rollback to 3.1

[07-Sep-2011 11:17:18] <rmatte> konvert: there are never rollback docs

[07-Sep-2011 11:17:26] <dhopp> konvert: do you have a backup?

[07-Sep-2011 11:17:36] <rmatte> konvert: you can do a hail mary by trying to install 3.1 overtop of what you have

[07-Sep-2011 11:17:41] <rmatte> no idea if it'll work or not though

[07-Sep-2011 11:18:14] <rmatte> konvert: you should really consider setting up a lab zenoss box to test upgrades before you roll them in to production

[07-Sep-2011 11:18:21] <rmatte> that way you don't get stuck in these situations

[07-Sep-2011 11:18:34] <rmatte> all my production stuff is still running 2.5.2 since 3.x still hasn't met my standards

[07-Sep-2011 11:19:09] <nexex> I had a snapshot of before the upgrade just in case, but uninstalling 3.2 rpms and installing the 3.1 worked for me

[07-Sep-2011 11:19:31] <rmatte> nexex: how did you save your data though?

[07-Sep-2011 11:21:01] <nexex> zenbackup

[07-Sep-2011 11:22:05] <rmatte> did you zenbackup while you were 3.1 or while you were 3.2?

[07-Sep-2011 11:22:12] <nexex> 3.1

[07-Sep-2011 11:22:25] <rmatte> right, so if konvert didn't then that method is useless

[07-Sep-2011 11:22:31] <konvert> so with zenbackup take backup of data, take snapshot, try to install 3.1 over 3.2, if it doesn't work, uninstall 3.2 and install 3.1 and reload data

[07-Sep-2011 11:22:45] <rmatte> konvert: correct, if you have a zenbackup from 3.1

[07-Sep-2011 11:22:52] <rmatte> I wouldn't try using a zenbackup from 3.2 on 3.1

[07-Sep-2011 11:22:59] <rmatte> it might work, but probably not

[07-Sep-2011 11:23:15] <rmatte> If you have a snapshot why don't you just roll back to that?

[07-Sep-2011 11:23:19] <rmatte> that's the method I use

[07-Sep-2011 11:24:00] <konvert> i have a really old snapshot form 2 months ago, i have done much work after it, but i think i have a zenbackup before upgrade

[07-Sep-2011 11:24:28] <Hackman238> rmatte: I'm disappointed to see these serious snmp bugs in Zenoss 3.2

[07-Sep-2011 11:24:41] <rmatte> konvert: ah, you should have done a snapshot right before the upgrade, rule of thumb

[07-Sep-2011 11:24:55] <konvert> i know, will do it in the future

[07-Sep-2011 11:25:07] <rmatte> Hackman238: yeh, clearly Zenoss QA isn't doing proper testing or they would have caught these, and it's not just snmp datasources, it's command datasources too

[07-Sep-2011 11:25:18] <Hackman238> rmatte: Madness!

[07-Sep-2011 11:25:32] <Hackman238> psyklone: Whos in charge of QA over there?

[07-Sep-2011 11:25:33] <konvert> but can i go back to the snapshot from two months ago and then load my zenbackup from just before upgrade

[07-Sep-2011 11:25:36] <rmatte> Hackman238: we're almost going to have to demand that ZCA perform testing and sign off on every release going forward to make sure that glaringly obvious bugs like these don't make it to the public

[07-Sep-2011 11:25:56] <Hackman238> rmatte: Its a shame, but looks like its needed.

[07-Sep-2011 11:26:06] <rmatte> Hackman238: this isn't the first time this has happened, it's almost at the point where the public cringes and cautiously upgrades every time there's a new version because you never know what's going to happen

[07-Sep-2011 11:26:21] <dhopp> Hackman238:  in my opinion the command datasources is a bigger issue..for the snmp you can just change the oid most of the time and get it working again…for command it requires code changes

[07-Sep-2011 11:26:22] <Hackman238> rmatte: I almost wonder if its intentional to a degree

[07-Sep-2011 11:26:25] <rmatte> you basically just hold your breath and run the install

[07-Sep-2011 11:26:34] <rmatte> I would hope not

[07-Sep-2011 11:26:42] <rmatte> let's not make assumptions like that lol

[07-Sep-2011 11:26:52] <konvert> but can i go back to the snapshot from two months ago and then load my zenbackup from just before upgrade

[07-Sep-2011 11:27:13] <rmatte> konvert: If you have a zenbackup that was taken while the system was 3.1, yes you can

[07-Sep-2011 11:27:57] <konvert> but where to chose to use the backup?

[07-Sep-2011 11:28:25] <dhopp> Hackman238:  I don't think it's intentional..but they released Enterprise and are trying to pull features out to make Core.  It's much harder to pull features rather then releasing core and adding features to make Enterprise

[07-Sep-2011 11:28:29] <Hackman238> dhopp: I didnt hear the details on the command datasource problem- whats the news with that?

[07-Sep-2011 11:28:30] <rmatte> you literally use the zenrestore command

[07-Sep-2011 11:28:51] <rmatte> Hackman238: It basically just doesn't parse some of the datapoints under the datasource

[07-Sep-2011 11:29:00] <dhopp> Hackman238:  basically if you have a command datasource (ApacheMonitor for example) that has multiple datapoints it will update the first datapoint and then exit

[07-Sep-2011 11:29:11] <Hackman238> dhopp: Thats the thing though....the Ent version of v3 has hardly any extra intergrated code to pull. Its mostly just extra addons.

[07-Sep-2011 11:29:14] <rmatte> yeh, that's brutality

[07-Sep-2011 11:29:29] <Hackman238> rmatte/dhopp: ah

[07-Sep-2011 11:29:50] <dhopp> Hackman238:  another community member found that reverting ZenRRD/parsers/Auto.py to the 3.1 version fixes that problem..but we don't know if that causes any other problems...

[07-Sep-2011 11:29:53] <konvert> i have done the procedure for a complete backup before upgrade to 3.2, can i do something with the complete-backup.tar.gz file?

[07-Sep-2011 11:30:04] <rmatte> konvert: zenrestore, read up on it

[07-Sep-2011 11:30:24] <rmatte> konvert: docs/DOC-3048

[07-Sep-2011 11:31:03] <konvert> thx

[07-Sep-2011 11:31:59] <dhopp> Hackman238:  based on the output of zencommand -v10 it gets all of the datapoints..just doesn't parse past the first one

[07-Sep-2011 11:32:39] <Hackman238> dhopp: Yeah sounds like if revertying Auto.py fixes it, maybe picking the parse format manually in the datasource might fix the problem

[07-Sep-2011 11:32:49] <Hackman238> *reverting

[07-Sep-2011 11:33:00] <psyklone> it will, changing it to nagios should do it as a temp workaround

[07-Sep-2011 11:34:04] <Hackman238> psyklone: Good to know. Thanks!

[07-Sep-2011 11:34:43] <rocket> rmatte/Hackman238: this issue your discussing is not an intentional regression,  I reported an issue with Auto.py earlier, though I dont recall what it was

[07-Sep-2011 11:35:08] <rocket> Hackman238: have you filed a bug report?

[07-Sep-2011 11:35:15] <Hackman238> rocket: Hey! How goes it? Thanks rocket.

[07-Sep-2011 11:35:19] <psyklone> rocket - both of these two issues are in as sev1's right now

[07-Sep-2011 11:35:45] <psyklone> i put them in yesterday after dhopp fired a flare

[07-Sep-2011 11:35:45] <Hackman238> rocket: Not yet, I think they are already filed though

[07-Sep-2011 11:36:03] <rocket> I believe the regex was too limited in the original Auto.py and it missed an edge case

[07-Sep-2011 11:36:04] <Hackman238> psyklone: Was a flare or cannon ball?

[07-Sep-2011 11:36:12] <Hackman238> rocket: Ah

[07-Sep-2011 11:36:20] <psyklone> Hackman238, it was a combination

[07-Sep-2011 11:36:20] <rocket> unfortunately updating regexes often comes with other additional side effects

[07-Sep-2011 11:36:40] <Hackman238> Appreciate all the input and work on the problem.

[07-Sep-2011 11:37:25] <Hackman238> rocket/psyklone: Starting NS Zen migrations today

[07-Sep-2011 11:37:32] <rocket> there is alot of work behinds the scenes on edge cases that you guys never see ...

[07-Sep-2011 11:37:50] <Hackman238> v2.5.2 + custom madness -> v4.0.2

[07-Sep-2011 11:38:24] <Hackman238> rocket: Absolutely, you do a great job.

[07-Sep-2011 11:38:37] <Hackman238> rocket: I trust you, I dont trust your management.

[07-Sep-2011 11:39:41] <dhopp> Hackman238:  I opened two tickets over the weekend..one for the oid issue and one for command issue..then was in here and a couple of community members were asking why their graphs were broken after an upgrade…klone messaged me and asked if it was post upgrade and I mentioned it was on a fresh install and he went running with it…I'm sure a fix is coming

[07-Sep-2011 11:40:00] <Hackman238> dhopp: Nice

[07-Sep-2011 11:40:31] <psyklone> dhopp - yes, fix is coming. thx for the heads up on those

[07-Sep-2011 11:41:48] <dhopp> rocket/psyklone: as a side note..front page of community.zenoss.org says 3.2.0 is now available but /forums says 3.1.0 is now available..heh..I haven't seen nyeates or I would give him hell :-P

[07-Sep-2011 11:42:09] <psyklone> dhopp - well i guess technically, both are true

[07-Sep-2011 11:42:12] <psyklone>

[07-Sep-2011 11:42:17] <dhopp> lol

[07-Sep-2011 11:42:49] <Hackman238> LOL

[07-Sep-2011 11:43:56] <psyklone> i think nick will have to fix that

[07-Sep-2011 11:47:38] <SDuensin> For the love of god or whoever you pray to, someone make psexec actually work!     (That is my rant for the day, thanks.)

[07-Sep-2011 11:50:08] <jb> hrm i need to write a nagios plugin, I guess.

[07-Sep-2011 11:50:37] <jb> actually, can I use a COMMAND datasource that results in a single integer, and use that as a datasource?

[07-Sep-2011 11:50:44] <jb> without doing a nagios-ish plugin?

[07-Sep-2011 11:51:16] <SDuensin> Hackman238, LOL

[07-Sep-2011 11:51:30] <SDuensin> But, no, really.  HELP!!!!

[07-Sep-2011 11:51:40] <Thinh> dont need nagios

[07-Sep-2011 11:51:49] <jb> sweet.  thanks!

[07-Sep-2011 11:51:56] <Thinh> i use the cacti format

[07-Sep-2011 11:52:05] <Thinh> var1:value1

[07-Sep-2011 11:52:40] <psyklone> jb - yes, if your command is just returning an integer you should be able to do that. set the parser format to nagios, and if your integer being returned is simply '5' that will get passed through even though it's not really being parsed out

[07-Sep-2011 11:52:46] <psyklone> more of a pass-through

[07-Sep-2011 11:53:11] <jb> ok, i'll give that a try.  thanks!

[07-Sep-2011 11:57:52] <rmatte> psyklone/rocket: it might be an idea for Zenoss to either take 3.2 down as a download or put a nice big warning above the download so that people are aware of these problems

[07-Sep-2011 11:58:14] <rmatte> a lot of people don't test before upgrading their production systems and then come in here and whine

[07-Sep-2011 11:58:58] <rmatte> the issues are basically show stoppers anyways

[07-Sep-2011 11:59:02] <dhopp> rmatte:  "I don't always test my code.  But when I do, I do it in production"

[07-Sep-2011 11:59:11] <Thinh> OLD!

[07-Sep-2011 11:59:12] <rmatte> dhopp: lol

[07-Sep-2011 11:59:18] <dhopp> Thinh:  but still funny

[07-Sep-2011 11:59:19] <Thinh> i have that picture printed

[07-Sep-2011 11:59:25] <dhopp> Thinh:  so do we

[07-Sep-2011 11:59:26] <dhopp> heh

[07-Sep-2011 11:59:31] <Thinh> i hand them out to QA when they cause an outage

[07-Sep-2011 12:00:01] [disconnected at Wed Sep  7 12:00:01 2011]

[07-Sep-2011 12:00:02] [connected at Wed Sep  7 12:00:02 2011]

[07-Sep-2011 12:00:19] [zenoss-logger (logger bot) has joined #zenoss]

[07-Sep-2011 12:00:37] <rocket> I feel for people who dont test, but as a former sysadmin I dont feel much

[07-Sep-2011 12:00:57] <rocket> whoever has the ticket opened just stress the urgency there

[07-Sep-2011 12:00:58] <dhopp> Hackman238:  oh come on..admit you upgrade production at Rack Space all the time without testing :-P

[07-Sep-2011 12:01:10] <Sam-I-Am> moo folks

[07-Sep-2011 12:01:17] <rocket> my internet connection isnt good enough today to do much from that standpoint :/

[07-Sep-2011 12:01:33] <twm1010> It never fails... the day after I finish building a Zenoss VM template, a new version comes out...

[07-Sep-2011 12:01:34] <willwh> rocket carrier pidgeon?

[07-Sep-2011 12:01:43] <dhopp> brb

[07-Sep-2011 12:01:56] <rocket> as Hackman238 testing first does lead to better results .. he can compare 4.0.2 to avalon beta for instance

[07-Sep-2011 12:02:08] <rocket> coffee shop .. my cable is out

[07-Sep-2011 12:02:12] <Sam-I-Am> lol

[07-Sep-2011 12:02:16] <Sam-I-Am> needs more intarnets

[07-Sep-2011 12:02:19] <twm1010> I think I'm going to start maintaining two templates, a bare Linux template, before Zenoss is installed, and one after.

[07-Sep-2011 12:02:47] <twm1010> Has anyone tried it on CentOS6 yet?

[07-Sep-2011 12:02:50] <dhopp> twm1010:  don't worry about upgrading to 3.2 right now

[07-Sep-2011 12:03:22] <rocket> twm1010: it might work, but it isnt tested internally yet

[07-Sep-2011 12:04:30] <rocket> resources are really tight until we can get off the 3.X code base and have a single 4.X release out there

[07-Sep-2011 12:04:38] <Hackman238> rocket: Oh yeah...I've done so many hundreeds of avalon tests I have an auto-rollout script

[07-Sep-2011 12:04:42] <rmatte> twm1010: you don't want this new version lol

[07-Sep-2011 12:05:01] <twm1010> i learned a long time ago not to be on the bleeding edge

[07-Sep-2011 12:05:26] <Hackman238> First thing I do everyday - su root; yum update --skip-broken -y

[07-Sep-2011 12:05:28] <rmatte> twm1010: yeh, I'm very much on the trailing edge when it comes to Zenoss

[07-Sep-2011 12:05:30] <Hackman238>

[07-Sep-2011 12:05:34] <rmatte> you learn after 2 versions

[07-Sep-2011 12:05:46] <jb> ugh im out of the zenoss loop.

[07-Sep-2011 12:05:46] <twm1010> I suppose I could save myself some time and script out my usual changes through the dmd

[07-Sep-2011 12:05:50] <jb> whats the newest enterprise version?

[07-Sep-2011 12:05:56] <Hackman238> jb: 4.0.2

[07-Sep-2011 12:06:02] <jb> ?!

[07-Sep-2011 12:06:04] <jb> stable?

[07-Sep-2011 12:06:08] <Hackman238> jb: Meh

[07-Sep-2011 12:06:10] <rmatte> yes

[07-Sep-2011 12:06:11] <Hackman238> jb: Mostly

[07-Sep-2011 12:06:14] <Hackman238> LOL

[07-Sep-2011 12:06:21] <twm1010> Theres about 30 minutes of things I put in after a clean install, regexes, other classes, etc, someday i'll stop doing it by hand

[07-Sep-2011 12:06:23] <jb> we just upgraded to 3 like 6 months ago.

[07-Sep-2011 12:06:42] <jb> i havent gotten any emails about 4

[07-Sep-2011 12:06:46] <Hackman238> jb: They've done a great job with 4.X. It's a huge move from v3

[07-Sep-2011 12:06:56] <Hackman238> jb: You run Ent?

[07-Sep-2011 12:06:58] <jb> yes

[07-Sep-2011 12:07:08] <SDuensin> Hackman238, any idea when us pleebs get to use 4.x?

[07-Sep-2011 12:07:08] <Hackman238> jb: Find your Zenoss Acct manager and slap him/her

[07-Sep-2011 12:07:14] <rmatte> lol

[07-Sep-2011 12:07:19] <jb> Hackman238: doing that now.

[07-Sep-2011 12:07:25] <Hackman238> SDuensin: Take a look at the ZCA timeline I put together

[07-Sep-2011 12:07:26] <twm1010> i wouldn't mind slapping her... she's got huge tracts of land

[07-Sep-2011 12:07:34] <twm1010> HUGE

[07-Sep-2011 12:07:46] <Hackman238> SDuensin: community/zca/blog/2011/08/30/zca-update-8-30-2011-operations-roadmap-announced

[07-Sep-2011 12:07:50] <SDuensin> Hackman238, ok...   Must have missed that while stuck in Windows Hell.

[07-Sep-2011 12:08:20] <Hackman238> twm1010: I dont understand the land comment.

[07-Sep-2011 12:08:25] <Hackman238> twm1010: Sorry. LOL

[07-Sep-2011 12:08:44] <twm1010> You need to brush up on your Monty Python

[07-Sep-2011 12:09:00] <Hackman238> twm1010: I know only Python, not Monty Python

[07-Sep-2011 12:09:05] <rmatte> which is actually what the name Python comes from, not the snake

[07-Sep-2011 12:09:06] <rmatte> lol

[07-Sep-2011 12:09:18] <Hackman238> rmatte: NO! LOL

[07-Sep-2011 12:09:30] <Hackman238> rmatte: No wonder Zope == madness

[07-Sep-2011 12:09:33] <twm1010> and I'm serious the account manager in chicago is pretty easy on the eyes

[07-Sep-2011 12:09:34] <Hackman238>

[07-Sep-2011 12:09:37] <rmatte> Hackman238: exactly

[07-Sep-2011 12:09:48] <Hackman238> twm1010: Oh gotcha.

[07-Sep-2011 12:09:49] <rmatte> twm1010: rock out

[07-Sep-2011 12:09:49] <jb> well, i don't know why the account manager in chicago hasn't contacted me.

[07-Sep-2011 12:09:49] <rmatte> lol

[07-Sep-2011 12:10:17] <jb> and what is the ZCA again?

[07-Sep-2011 12:10:24] <twm1010> community alliance

[07-Sep-2011 12:10:26] <Hackman238> jb: Zenoss-Community Alliance

[07-Sep-2011 12:10:34] <jb> and what is the new enterprise download portal URK?

[07-Sep-2011 12:10:36] <jb> URL?

[07-Sep-2011 12:10:38] <twm1010> it's a world of warcraft clan, hell bent on monitoring the orcish hordes

[07-Sep-2011 12:10:39] <jb> wtf

[07-Sep-2011 12:10:53] <Hackman238> jb: 1 sec

[07-Sep-2011 12:10:59] <Thinh> zenoss got some cool features but its too buggy

[07-Sep-2011 12:11:00] <rocket> support.zenoss.com

[07-Sep-2011 12:11:01] <twm1010> i'm in a mood today, can you tell?

[07-Sep-2011 12:11:04] <Thinh> and it cant handle high volume

[07-Sep-2011 12:11:07] <rmatte> twm1010: "threshold of Horde exceeded: run"

[07-Sep-2011 12:11:13] <Hackman238> jb: PM'd to you

[07-Sep-2011 12:11:23] <jb>

[07-Sep-2011 12:11:29] <jb> Invalid deptID in URL. Please make sure you have the entire, correct URL.

[07-Sep-2011 12:11:30] <jb> :/

[07-Sep-2011 12:11:33] <rocket> Thinh: those blanket statements are not entirely accurate

[07-Sep-2011 12:11:43] <Thinh> oh yea?

[07-Sep-2011 12:11:51] <rocket> Thinh: we have very high volume accounts on 4.X

[07-Sep-2011 12:11:57] <rmatte> Thinh: It can handle high volume if you tune it and set it up correctly

[07-Sep-2011 12:11:59] <rocket> I would agree 3.X can only go so far

[07-Sep-2011 12:12:00] <Thinh> heh

[07-Sep-2011 12:12:09] <Thinh> 4.x now

[07-Sep-2011 12:12:29] <twm1010> rocket: 4.x still uses RRD?

[07-Sep-2011 12:12:37] <rmatte> Thinh: and a lot of the bugs have to do with the new UI which is still young

[07-Sep-2011 12:12:40] <Hackman238> Thinh: 4.x brings 1000x improvement and 1000yrs of darkness

[07-Sep-2011 12:12:52] <jb> is there a list of changes/improvements in 4?

[07-Sep-2011 12:12:59] <rmatte> rocket: the word "Override" has to go, do whatever you have to

[07-Sep-2011 12:13:13] <Hackman238> rmatte: sed

[07-Sep-2011 12:13:33] <rmatte> sed 's/Override/Copy/g'

[07-Sep-2011 12:13:35] * Sam-I-Am needs to get a zenoss vm going here

[07-Sep-2011 12:13:40] <Sam-I-Am> getting rusty :/

[07-Sep-2011 12:13:45] <Thinh> thing is i dont like seeing so many upgrades when each version is still full of bugs and the upgrade process break a bunch of stuff heh

[07-Sep-2011 12:13:53] <rocket> twm1010: it does but it also uses the rrd caching daemon

[07-Sep-2011 12:14:05] <Sam-I-Am> <3 rrdcached

[07-Sep-2011 12:14:06] <twm1010> cool

[07-Sep-2011 12:14:10] <Sam-I-Am> that must make a huge improvement

[07-Sep-2011 12:14:10] <hmp> 4.x should be da shit

[07-Sep-2011 12:14:10] <rmatte> Thinh: yeh well, welcome to Zenoss lol

[07-Sep-2011 12:14:24] <hmp> i just hope it upgrades from 3.x easily

[07-Sep-2011 12:14:36] <jb> is there a list of changes/improvements in 4?

[07-Sep-2011 12:14:38] <Thinh> my friend told me they have a dedicated zenoss engineer trying to get their shit working right

[07-Sep-2011 12:14:46] <Thinh> they migrated to oracle and its still crapping out

[07-Sep-2011 12:14:58] <rocket> migrated what to oracle?

[07-Sep-2011 12:15:01] <Hackman238> Thinh: Zenoss brings a lot of functionality that would be painful or impossible to do in a timely fashion with Nagios, Cati and others

[07-Sep-2011 12:15:04] <Thinh> mysql to oracle

[07-Sep-2011 12:15:08] <Simon4> woah

[07-Sep-2011 12:15:09] <jb> 4 runs oracle?

[07-Sep-2011 12:15:13] <rocket> no

[07-Sep-2011 12:15:13] <jb> or is that a custom case?

[07-Sep-2011 12:15:15] <hmp> events db is rarely an issue

[07-Sep-2011 12:15:17] <Thinh> custom yo

[07-Sep-2011 12:15:17] <Simon4> that's one way to try and fix a problem

[07-Sep-2011 12:15:22] <rocket> it does not and I know of no customer doing that

[07-Sep-2011 12:15:24] <Hackman238> Thinh: but this same functionality also brings the possibility of complication here and there

[07-Sep-2011 12:15:28] <SDuensin> Oooo - new lead!  Anyone use kerberos in Linux against a Windows domain?

[07-Sep-2011 12:15:34] <Hackman238> jb: 4.x runs on mysql 5.5

[07-Sep-2011 12:15:37] <Sam-I-Am> rocket: where were you in MN again?

[07-Sep-2011 12:15:38] <jb> ok

[07-Sep-2011 12:15:44] <rocket> north of the twin cities

[07-Sep-2011 12:15:49] <jb> Hackman238: is there anywhere where I can ready about updates/differences in 4.x?

[07-Sep-2011 12:15:50] <Hackman238> SDuensin: Negative, dodgey in my exp.

[07-Sep-2011 12:16:02] <Hackman238> jb: Let me take a look

[07-Sep-2011 12:16:02] <SDuensin> Hackman238, and what with Windows isn't?!

[07-Sep-2011 12:16:03] <rocket> I am a one off former community member ... a rarity ..

[07-Sep-2011 12:16:05] <jb> Hackman238: read, sorry.

[07-Sep-2011 12:16:09] <Sam-I-Am> ah, i have a client in minneapolis... maybe sometime when i'm up there we can meet

[07-Sep-2011 12:16:17] <rmatte> SDuensin: well, I have successfully set that up in the past for a customer, but I don't personally do it with our stuff

[07-Sep-2011 12:16:21] <dhopp> SDuensin:  I have in the past..but not currently...

[07-Sep-2011 12:16:25] <twm1010> anyone in the midwest/chicagoland ?

[07-Sep-2011 12:16:26] <Thinh> dont get me wrong, i think zenoss is great but its terrible when you have reach high volume =0

[07-Sep-2011 12:16:36] <dhopp> twm1010:  I am

[07-Sep-2011 12:16:45] <rmatte> Thinh: well, define high volime

[07-Sep-2011 12:16:48] <rmatte> volume*

[07-Sep-2011 12:16:49] <dhopp> Thinh:  define "high volume"

[07-Sep-2011 12:16:52] <Thinh> sure

[07-Sep-2011 12:16:52] <dhopp> rmatte:  lol

[07-Sep-2011 12:16:52] <rocket> Sam-I-Am: sure

[07-Sep-2011 12:17:02] <Thinh> what you need to know

[07-Sep-2011 12:17:10] <SDuensin> rmatte, dhopp - "all I want it to do is..." allow me to perform an implicit login using credentials in the cache.  Does that make sense?

[07-Sep-2011 12:17:23] <twm1010> I was able to get a much older server to monitor 600 devices, with a TON of datapoints, with some basic tuning....

[07-Sep-2011 12:17:29] <Hackman238> jb: There must be somewhere, but I dont have one. Sorry

[07-Sep-2011 12:17:31] <rocket> Thinh: in 4.X we have the event system handling 8million events in 4 days ..

[07-Sep-2011 12:17:37] <twm1010> Where I found it inflexible is a need to have a different data collection cycle for different things

[07-Sep-2011 12:17:37] <jb> Hackman238: k.

[07-Sep-2011 12:17:58] <rmatte> twm1010: yeh, the data collection cycle limitation constantly comes back and bites me in the ass

[07-Sep-2011 12:17:59] <rmatte> lol

[07-Sep-2011 12:18:09] <Hackman238> rocket: Sounds like our syslog post filtering *chuckles*

[07-Sep-2011 12:18:31] <rmatte> we'll have a customer say "We want all interfaces monitored at a minute but everything else at 5", sorry, no can do

[07-Sep-2011 12:18:43] <Hackman238> rocket: Fingers crossed that we'll be able to get rid of splunk and use only Zenoss!

[07-Sep-2011 12:19:06] <Thinh> rocket, im looking forward to testing 4x

[07-Sep-2011 12:19:14] <Hackman238> twm1010: You only need to add logical collectors for varied rates

[07-Sep-2011 12:19:18] <rmatte> Thinh: when you define "high volume" how many datasources are you talking about?

[07-Sep-2011 12:19:43] <rmatte> I'm doing over 20k datasources on one of my servers and the polling cycles takes less than a minute each cycle

[07-Sep-2011 12:19:50] <rmatte> pretty efficient as far as I'm concerned

[07-Sep-2011 12:19:58] <rocket> snmp polling is the shiz in zenoss ..

[07-Sep-2011 12:20:03] <rmatte> and that's monitoring across VPN tunnels too

[07-Sep-2011 12:20:09] <twm1010> Hackman238: sure, i can do that.... but what if someone wants the server monitored at 5 minutes, but apache_mod_status statistics from the same server at 1 minute?

[07-Sep-2011 12:20:11] <twm1010> gets fun

[07-Sep-2011 12:20:23] <Hackman238> There is nothing I've seen that can doll snmp packets like twisted in Zenoss can.

[07-Sep-2011 12:20:26] <rocket> ssh polling is quite impressive in that most tools dont even do it .. or require special clients etc

[07-Sep-2011 12:20:31] <Thinh> rmatte at citrix we had 5,000 devicess as far as datasource its about 250+ for each server

[07-Sep-2011 12:20:31] <Sam-I-Am> rocket: now if only you could convince this client to go with zenoss

[07-Sep-2011 12:20:33] <rmatte> Hackman238: yeh, it's insane

[07-Sep-2011 12:20:49] <Thinh> they are all jmx and custom monitoring for goto product

[07-Sep-2011 12:20:59] <Thinh> i shouldnt care really sicne i dont work there anymore heh

[07-Sep-2011 12:21:02] <Hackman238> twm1010: Not at all, just add it on the 1 minute poller and set the datasource for the 5 minute items to 5 minutes, else to 1

[07-Sep-2011 12:21:08] <rmatte> Thinh: ok, you could still do that, you'd just need multiple collectors

[07-Sep-2011 12:21:09] <Hackman238> twm1010: BAM! 5min and 1 min

[07-Sep-2011 12:21:18] <Thinh> matte we had 12 zencollectors

[07-Sep-2011 12:21:29] <rmatte> Thinh: there aren't many monitoring tools on the planet that can do 1250000 datasources with a single collector (if any)

[07-Sep-2011 12:21:46] <Thinh> notlooking to use 1

[07-Sep-2011 12:21:56] <rmatte> you'd need more than 12

[07-Sep-2011 12:22:00] <twm1010> dhopp: where you located?

[07-Sep-2011 12:22:08] <Thinh> idea was to have 1 per datacenter

[07-Sep-2011 12:22:18] <rmatte> even with 12 that's still 104166 per collector

[07-Sep-2011 12:22:29] <rmatte> if you had like 20 collectors you'd probably be able to pull it off

[07-Sep-2011 12:22:34] <rmatte> and you'd need some serious hardware

[07-Sep-2011 12:22:37] <Hackman238> Thinh: Your disk will be a huge limit. You;ll need a RAID 10 of Fusion IO cards to meet that demand

[07-Sep-2011 12:22:42] <twm1010> rmatte: I'm about to get some resources running some ESX based zenoss boxes for monitoring hosted services

[07-Sep-2011 12:22:47] <rmatte> yeh, disk is everything

[07-Sep-2011 12:22:57] <dhopp> let's say I have a linux device that I monitor…the collector perf is cycle is at 5 minutes..can't I add a command datasource (check_http for example) that has a 1 minute cycle to that deive?

[07-Sep-2011 12:23:00] <dhopp> *device

[07-Sep-2011 12:23:05] <twm1010> i should say managed services

[07-Sep-2011 12:23:18] <rmatte> I migrated one of my VMs from a slower SAN to a new fast SAN that we got and it went from load averages in the double digits to under 1

[07-Sep-2011 12:23:22] <Hackman238> dhopp: No, but if you ahev a 1 minute poller you can add 5 minute datasources

[07-Sep-2011 12:23:27] <rocket> most often though small polling cycles usually dont make sense

[07-Sep-2011 12:23:53] <Hackman238> dhopp: 1 divides in to 5, 5 does not divide in to 5 in the context of time.

[07-Sep-2011 12:24:00] <dhopp> Hackman238:  what about for snmp?

[07-Sep-2011 12:24:12] <dhopp> what I mean is can't I have snmp at 5 minutes and command at 1 minute?

[07-Sep-2011 12:24:32] <rmatte> dhopp: you can increase the zencommand cycle time to 1 minute and then set the command datasource to a minute, and it'll run every minute while the snmp stuff will still run at 5, does that answer your question?

[07-Sep-2011 12:24:34] <Hackman238> dhopp: Sure you can, the snmp daemon is seperate

[07-Sep-2011 12:24:44] <dhopp> that's what I thought

[07-Sep-2011 12:24:45] <dhopp> ok

[07-Sep-2011 12:24:55] <Hackman238> dhopp: You'll need to make sure all the RRD create commands are right for the cycle times, however

[07-Sep-2011 12:25:00] <dhopp> rmatte/Hackman238: thanks

[07-Sep-2011 12:25:04] <dhopp> shit..br b

[07-Sep-2011 12:25:08] <Hackman238> dhopp: NP

[07-Sep-2011 12:25:35] <rmatte> yeh, the RRD create commands need to be configured on zencommand with a lower cycle time, and you may have to actually configure them in the monitoring template itself if you want more granularity

[07-Sep-2011 12:25:39] <Hackman238> Thinh: 1250000dp on one collector isn't realistic

[07-Sep-2011 12:25:46] <rmatte> for a lower cycle time rather

[07-Sep-2011 12:26:02] <Hackman238> Thinh: If a row or cab goes down it'll up the cycle time and shitstorm will ensue

[07-Sep-2011 12:26:16] <rmatte> Hackman238: well, he wanted it spread across 12

[07-Sep-2011 12:26:26] <Hackman238> rmatte: Oh my bad

[07-Sep-2011 12:26:27] <rmatte> Hackman238: but even then, that's a lot, it's 104166 per collector

[07-Sep-2011 12:26:35] <Thinh> rocket knows =)

[07-Sep-2011 12:26:35] <rmatte> he'd need some serious IO and hardware

[07-Sep-2011 12:26:48] <Hackman238> rmatte: He'll need SSD cards, yes

[07-Sep-2011 12:26:54] <Hackman238> rmatte: Otherwise it'll never work.

[07-Sep-2011 12:26:57] <rmatte> yeh

[07-Sep-2011 12:26:59] <Thinh> you have no idea what citrix can afford =)

[07-Sep-2011 12:27:04] <Hackman238> rmatte: Just add pwoer and instant shit storm

[07-Sep-2011 12:27:05] <Thinh> we had ssd for all

[07-Sep-2011 12:27:17] <jb> ok, I added a COMMAND datasource, tested it successfuly against a device, but it's not appearing in the "Add Data Point" dialog when I create a graph.

[07-Sep-2011 12:27:26] <rmatte> Thinh: there are also several tuning options in Zenoss that need to be changed to optimize it

[07-Sep-2011 12:27:33] <rmatte> Thinh: stock Zenoss Core is tuned to be slow

[07-Sep-2011 12:27:50] <Thinh> we ran ent with all the tuning from zenoss support

[07-Sep-2011 12:27:51] <Hackman238> Thinh: Do you use Zenoss?

[07-Sep-2011 12:28:03] <Thinh> not here at the moment hackman

[07-Sep-2011 12:28:07] <Thinh> opennms at my current job

[07-Sep-2011 12:28:09] <rmatte> Thinh: then provided the hardware was adequate it should have worked

[07-Sep-2011 12:28:33] <rmatte> OpenNMS is better for high load since it was written with that sort of thing in mind from the start

[07-Sep-2011 12:28:40] <jb> hrm, wtf.

[07-Sep-2011 12:28:42] <Hackman238> Thinh: gotcha. I know of nothing opensource or opensource-ish that will do what core will with a few tweaks

[07-Sep-2011 12:28:44] <rmatte> it's also not running on a scripting language

[07-Sep-2011 12:29:09] <rmatte> yeh, OpenNMS make have good performance but it's lacking in features compared to Zenoss

[07-Sep-2011 12:29:14] <rmatte> may have*

[07-Sep-2011 12:29:15] <Thinh> right

[07-Sep-2011 12:29:18] <Thinh> also the GUI sux

[07-Sep-2011 12:29:28] <jb> any ideas?

[07-Sep-2011 12:29:35] <rmatte> yeh, I took a look at it a while back and hated the GUI

[07-Sep-2011 12:29:47] <Hackman238> Thinh: There's a balance of functionality, custimization and programmatic manageability not found elsewhere

[07-Sep-2011 12:30:08] <Hackman238> jb: try F5

[07-Sep-2011 12:30:21] <Hackman238> jb: Chances are AJAX failed to notify correctly

[07-Sep-2011 12:30:21] <Thinh> heh thats like the counterstrike trick

[07-Sep-2011 12:30:23] <jb> i did that

[07-Sep-2011 12:30:23] <Thinh> f10!

[07-Sep-2011 12:30:23] <rmatte> Hackman238: yeh, for integration with other systems there really is not better alternative

[07-Sep-2011 12:30:37] <Hackman238> jb: Hum. Thats odd.

[07-Sep-2011 12:30:38] <jb> i hate the buggy ajax problems in 3.x, btw.

[07-Sep-2011 12:30:44] <jb> the event console is UGH.

[07-Sep-2011 12:30:54] <Hackman238> jb: Its all fixed in v4

[07-Sep-2011 12:31:00] <jb> thank god!

[07-Sep-2011 12:31:09] <Hackman238> jb: It works so well it'll make you love AJAX

[07-Sep-2011 12:31:10] <jb> is the UI much different?

[07-Sep-2011 12:31:14] <Hackman238> jb: Nope.

[07-Sep-2011 12:31:16] <rmatte> Hackman238: whenever the hell that actually gets released to core lol

[07-Sep-2011 12:31:23] <Hackman238> jb: Just coded better

[07-Sep-2011 12:31:32] <Hackman238> rmatte: Soon!

[07-Sep-2011 12:31:45] <jb> well this is annoying, I really want this graph

[07-Sep-2011 12:31:53] <rmatte> Hackman238: I've been dying for a update that doesn't suck for ages now, it's been over a year and still nothing

[07-Sep-2011 12:32:17] <jb> command template is "/opt/zenoss/libexec/ssltps.sh "${dev/manageIp}""

[07-Sep-2011 12:32:21] <jb> "TEST" works..

[07-Sep-2011 12:32:23] <Hackman238> rmatte: Dont give up! The summit is just a little further

[07-Sep-2011 12:32:25] <Thinh> i also like to be able to put a date range on my graph instead of clicking the back and forth button =0

[07-Sep-2011 12:32:35] <rmatte> Hackman238: yeh, I'm hoping so, very anxious

[07-Sep-2011 12:32:45] <Hackman238> Thinh: You can, thats a very easy pack

[07-Sep-2011 12:32:45] <jb> Thinh: I suggested that to E. Dahl about 4 years ago.

[07-Sep-2011 12:33:05] <Thinh> easy pack eh

[07-Sep-2011 12:33:09] <rmatte> Hackman238: that should be a built in feature really

[07-Sep-2011 12:33:17] <rmatte> selecting date ranges is common functionality

[07-Sep-2011 12:33:19] <Hackman238> Tell you what- We'll release the Rackspace graph range UI pack addon

[07-Sep-2011 12:33:32] <jb> PLEASE do.

[07-Sep-2011 12:33:34] <rmatte> Hackman238: do you have one for 2.5?

[07-Sep-2011 12:33:38] <jb> i really want graph ranges.

[07-Sep-2011 12:33:43] <Hackman238> rmatte: Its only for 2.5.2 and 4.0

[07-Sep-2011 12:33:49] <Thinh> pffft

[07-Sep-2011 12:33:52] <Hackman238> rmatte: We'll need to port or backport to v3

[07-Sep-2011 12:33:55] <rmatte> Hackman238: excellent, I'd love to get my hands on the 2.5.2 one

[07-Sep-2011 12:33:58] <jb> i have been requesting real graph ranges for years :/

[07-Sep-2011 12:34:11] <rmatte> Hackman238: maybe we can hack together support for 3.x

[07-Sep-2011 12:34:13] <Hackman238> rmatte: Do me a favor, I'll send it to you, but try to migrate to v3 for the others

[07-Sep-2011 12:34:24] <rmatte> Hackman238: yup, will do

[07-Sep-2011 12:34:25] <Hackman238> rmatte: I dont have time right now to migrate it myself

[07-Sep-2011 12:34:32] <Hackman238> rmatte: I'll get to you in a day or so.

[07-Sep-2011 12:34:40] <rmatte> no rush

[07-Sep-2011 12:34:45] <Hackman238> Thinh: Sorry, we dont use v3

[07-Sep-2011 12:35:01] <rmatte> A lot of people are still no 2.5.2 since it's stable as hell

[07-Sep-2011 12:35:04] <jb> sweet.. can't wait to see the graph plugin.

[07-Sep-2011 12:35:07] <rmatte> best release I've seen since I started using Zenoss

[07-Sep-2011 12:35:14] <rmatte> on*

[07-Sep-2011 12:35:14] <jb> yeah i wish we wouldn't have upgraded to 3.x.

[07-Sep-2011 12:35:17] <jb> too many bugs

[07-Sep-2011 12:35:20] <Thinh> ^

[07-Sep-2011 12:35:51] <jb> damnit, this datasource problem is not going away.

[07-Sep-2011 12:36:03] <Thinh> whats hte problem

[07-Sep-2011 12:36:05] <rmatte> jb: have you tried staring at it harder?

[07-Sep-2011 12:36:10] <jb> yes!

[07-Sep-2011 12:36:14] <rmatte> damn

[07-Sep-2011 12:36:18] <rmatte> usually works for me

[07-Sep-2011 12:36:25] <jb> I created a COMMAND datasource, but it's not showing up as an avaibale graph point.

[07-Sep-2011 12:36:33] <Thinh> did you add the datasource first

[07-Sep-2011 12:36:38] <jb> ?

[07-Sep-2011 12:36:45] <rmatte> alrighty, afk for a few, have to add some devices to *cringe* solarwinds

[07-Sep-2011 12:37:05] <Thinh> did you check if there is an rrd file

[07-Sep-2011 12:38:00] <jb> oh

[07-Sep-2011 12:38:06] <jb> another crappy UI problem

[07-Sep-2011 12:38:31] <Hackman238> http://shanewilliamscott.com/img/date.png

[07-Sep-2011 12:38:44] <jb> Hackman238: thats sweet.

[07-Sep-2011 12:39:36] <rmatte> Hackman238: I doubt that'd be hard to duplicate in 3.x

[07-Sep-2011 12:39:43] <Hackman238> rmatte: Nope, its not.

[07-Sep-2011 12:39:44] <rmatte> I'll take a stab at it when you pass me the pack

[07-Sep-2011 12:39:50] <Hackman238> rmatte: TY

[07-Sep-2011 12:41:43] <jb> ok, works now.

[07-Sep-2011 12:41:58] <Hackman238> jb: Nice

[07-Sep-2011 12:42:12] <jb> couldn't figure out a way to poll multiple SNMP OID's and then do calculations on them.. so I just used a COMMAND datasource

[07-Sep-2011 12:42:30] <jb> Hackman238: you guys have BIGIP's at RS?

[07-Sep-2011 12:42:51] <Hackman238> jb: *face palm* yes

[07-Sep-2011 12:43:19] <jb> doing anything cool/custom in terms of monitoring?

[07-Sep-2011 12:43:27] <jb> the stock enterprise zenpack leaves a little to be desired.

[07-Sep-2011 12:43:38] <Hackman238> jb: We're building our own pack since the ent one sucks

[07-Sep-2011 12:43:43] <jb> man i know.

[07-Sep-2011 12:43:44] <Hackman238> jb: Agreed

[07-Sep-2011 12:43:54] <jb> ive been beefing it up the past few days

[07-Sep-2011 12:43:57] <jb> but i still don't like it.

[07-Sep-2011 12:43:57] <Hackman238> jb: We'll release it to the coomunity once done

[07-Sep-2011 12:44:02] <jb> can't wait.

[07-Sep-2011 12:44:10] <jb> the COMMAND datasource I just wrote monitors SSL TPS.

[07-Sep-2011 12:44:14] <Hackman238> jb: We have ASR 7/9K pack needing testing

[07-Sep-2011 12:44:22] <jb> hm, no ASR here.

[07-Sep-2011 12:44:24] <Hackman238> jb: Gotcha

[07-Sep-2011 12:44:26] <jb> we are still 6500 based.

[07-Sep-2011 12:44:34] <Hackman238> jb: Ouch. Where you at?

[07-Sep-2011 12:44:41] <jb> follett.com

[07-Sep-2011 12:45:12] <Hackman238> jb: The ASR 7/9K's are beasts. DoS it all you want- rolls off its back like work to an innercity stoop warmer.

[07-Sep-2011 12:45:13] <jb> Hackman238: do you guys have a repo or anything of your zenpacks that you may have released?

[07-Sep-2011 12:45:24] <jb> Hackman238: yeah, but you guys do about 1000x the traffic that we do

[07-Sep-2011 12:45:38] <Hackman238> jb: Soon the ZCA will be releasing a development timeline

[07-Sep-2011 12:45:42] <jb> our total internet transit is way under 1GB.

[07-Sep-2011 12:46:15] <jb> awesome.

[07-Sep-2011 12:46:28] <jb> im definetely interested in your bigip pack, and the graph range.

[07-Sep-2011 12:46:28] <Hackman238> jb: Oh wow yeah- we're up around 30x that

[07-Sep-2011 12:46:36] <jb> i thought so

[07-Sep-2011 12:46:43] <Hackman238> jb: Absolutely. Stay tuned!

[07-Sep-2011 12:46:46] <Hackman238> BRBB

[07-Sep-2011 12:46:49] <jb> also, let me know if you need any help/testing.

[07-Sep-2011 12:47:06] <jb> i have stacks of bigips

[07-Sep-2011 12:50:56] <dhopp> what is the default zencommand cycle time?

[07-Sep-2011 12:52:39] <Hackman238> jb: Will do!

[07-Sep-2011 12:52:49] <Hackman238> dhopp: 5 min I think

[07-Sep-2011 12:53:39] <dhopp> Hackman238:  where is that defined?  don't see anything in the ui

[07-Sep-2011 12:58:08] <JohnnyNOC> dhopp i think it's per collector

[07-Sep-2011 12:58:52] <JohnnyNOC> hrm, actually, i don't see anything about zencommand cycles

[07-Sep-2011 12:59:22] <JohnnyNOC> unless it falls under the 'config cycle interval'

[07-Sep-2011 13:04:49] <dhopp> oh this is awesome:

[07-Sep-2011 13:04:50] <dhopp> 2011-09-07 07:31:33,620 INFO zen.zencommand: config: <Products.ZenRRD.zencommand.DeviceConfig instance at 0x11c269e0>

[07-Sep-2011 13:05:02] <dhopp> what does that mean?  is that hex supposed to be the timestamp?

[07-Sep-2011 13:06:48] <dhopp> which wouldn't make sense since there is a timestamp in the log already

[07-Sep-2011 13:09:59] <luizzmizz> hi there!

[07-Sep-2011 13:11:35] <luizzmizz> i'm finishing my printermib zenpack and i'm completely stuck into the toners graphical interface... i get to show some info into the Details tab, but not on the toners (as a component of the printer device) panel

[07-Sep-2011 13:11:42] <luizzmizz> anyone could give me a hint?

[07-Sep-2011 13:12:03] <luizzmizz> i now it's almost finished but i don't see the error...

[07-Sep-2011 13:13:37] <luizzmizz> please?

[07-Sep-2011 13:18:48] <dhopp> luizzmizz:  I'm not sure (I haven't developed a zenpack that was more then just a collection of templates).  Have you posted anything on the forums?

[07-Sep-2011 13:21:02] <luizzmizz> dhopp: nope, i thought it was better to speak about it here :S

[07-Sep-2011 13:21:18] <dhopp> luizzmizz:  sometimes it is..kind of depends on who is hovering

[07-Sep-2011 13:21:41] <luizzmizz> well, meanwhile i don't find the person, i will follow your advice, thanx

[07-Sep-2011 13:21:41] <luizzmizz> !

[07-Sep-2011 13:21:46] <luizzmizz> (dhopp)

[07-Sep-2011 13:23:47] <dhopp> next thursday (week from tomorrow) I believe is a developer chat.  There will be a couple of zenoss inc. devs in here that might be able to help..

[07-Sep-2011 13:26:35] <luizzmizz> dhopp: good to know, thanks

[07-Sep-2011 13:26:58] <dhopp> I *think* it starts at 11AM ET..I can never remember the time

[07-Sep-2011 13:27:05] <dhopp> it usually goes for 60-90 minutes

[07-Sep-2011 13:27:16] <dhopp> kind of depends on what questions are being asked

[07-Sep-2011 13:41:37] <jb> hrm, anybody using bind9's stat channels to pull bind perf data?

[07-Sep-2011 13:43:01] <dhopp>  Hackman238/JohnnyNOC: I just did a quick test and bound Apache Template and HttpMonitor template to the same device.  Both have a default Cycle time of 300.  I changed HttpMonitor to have a cycle time of 60 and the RRDs get created with the correct step (60 for HttpMonitor and 300 for Apache).  Am I missing something obvious?

[07-Sep-2011 13:44:31] <jb> hrm, or can zenoss parse XML data natively?

[07-Sep-2011 13:44:37] <jb> as a datasource.

[07-Sep-2011 13:45:00] <dhopp> jb: good question..I don't know :-P

[07-Sep-2011 13:47:44] <twm1010> dhopp: interesting to see that behavior, maybe something changed at some point, i remember wrestling with it

[07-Sep-2011 13:48:14] <twm1010> i've been wanting to give myself a refresher in RRD to see about keeping details a while longer

[07-Sep-2011 13:48:22] <twm1010> on top of the polling cycle changes

[07-Sep-2011 13:53:06] <dhopp> twm1010:  and the rrd seems to get updated correctly.

[07-Sep-2011 13:53:40] <dhopp> twm1010:  maybe that's something that changed in 3.x?  I just want to make sure I'm not missing something obvious and then in 6 months have it blow up in my face

[07-Sep-2011 13:56:04] <twm1010> all i know is, in 2.5.x i had some issues with that, and based on what you've tested, that's awesome

[07-Sep-2011 13:56:40] <Gat0rvean> Is there a way to turn off performance graphs

[07-Sep-2011 13:56:47] <Gat0rvean> for an entire device class

[07-Sep-2011 13:56:53] <rmatte> Gat0rvean: unbind the template from the class

[07-Sep-2011 13:57:02] <jb> is there a trick to redirecting zenperfsnmp output to a file?

[07-Sep-2011 13:57:09] <Gat0rvean> rmatte: where is it bound?

[07-Sep-2011 13:57:35] <rmatte> Gat0rvean: This is covered in the Zenoss Admin guide just so you know, but it'll be under the "DETAILS" button in the Infrastructure section

[07-Sep-2011 13:57:45] <rmatte> just make sure you have the class highlighted on the left before you click the button

[07-Sep-2011 13:57:50] <rmatte> it's a big blue button shaped like an arrow

[07-Sep-2011 13:58:33] <jb> zenperfsnmp > file doesn't seem to be working.

[07-Sep-2011 13:58:51] <dhopp> Gat0rvean:  I'm not sure why you want to do that?  If you turn off the graphs (i.e. unbind the template) you won't get any alerting for those either.  So you will just have the device in the zenoss but it won't be doing anything

[07-Sep-2011 13:59:18] <rmatte> Gat0rvean: actually you don't even have to click details, just select the class, then in the left bottom menu, click on the menu with the gear icon and select bind templates

[07-Sep-2011 13:59:41] <rmatte> Gat0rvean: then just make sure there are no templates in the Selected section

[07-Sep-2011 13:59:51] <dhopp> jb: zenperfsnmp 2>&1> file ?

[07-Sep-2011 14:00:18] <Gat0rvean> dhopp: I just want no graphs, the blades send traps to zenoss that sends out alerts, are you saying if I unbind the templates it won't send alerts on my traps eitehr?

[07-Sep-2011 14:00:31] <rmatte> dhopp: it's technically > file 2>&1

[07-Sep-2011 14:00:38] <dhopp> Gat0rvean:  traps should still work

[07-Sep-2011 14:00:44] <dhopp> rmatte:  I think it works either way :-P

[07-Sep-2011 14:00:47] <rmatte> though I guess it works both ways

[07-Sep-2011 14:00:56] <jb> thanks.

[07-Sep-2011 14:00:56] <rmatte> I just always see the 2>&1 done after

[07-Sep-2011 14:01:03] <jb> ok, so no native XML datasource types in zen (yet).

[07-Sep-2011 14:01:06] <jb> maybe in 4?

[07-Sep-2011 14:01:34] <rmatte> doubt it, 4 is mostly about enhanced event management from what I hear

[07-Sep-2011 14:01:38] <dhopp> rmatte: I usually see it your way too..but I couldn't remember the exact syntax so I looked at a crontab of one of our servers and it had it the other way so that's what I copied and pasted :-P

[07-Sep-2011 14:01:58] <rmatte> ah ok

[07-Sep-2011 14:01:58] <Gat0rvean> rmatte:  your saying under Infrastructure > Network > MY CONTAINER > Details > Monitoring Templates: just remove the Device (/Devices) object?

[07-Sep-2011 14:02:36] <rmatte> Gat0rvean: no, to repeat what I said: "actually you don't even have to click details, just select the class, then in the left bottom menu, click on the menu with the gear icon and select bind templates"

[07-Sep-2011 14:03:09] <rmatte> "then just make sure there are no templates in the Selected section"

[07-Sep-2011 14:03:38] <Gat0rvean> rmatte: ahhhh, awesome, and this will only affect the class I have selected correct?

[07-Sep-2011 14:03:43] <rmatte> correct

[07-Sep-2011 14:04:09] <Gat0rvean> rmatte: thank you

[07-Sep-2011 14:04:37] <dhopp> Gat0rvean:  and every class / device below it

[07-Sep-2011 14:04:49] <rmatte> it gets set as a zProperty on the class, so if you ever need to verify it just check the "zDeviceTemplates" property

[07-Sep-2011 14:04:51] <rmatte> you're welcome

[07-Sep-2011 14:05:12] <rmatte> and as dhopp said, it'll affect everything under that class, even subclasses

[07-Sep-2011 14:05:20] <rmatte> unless you do the same thing to the sub classes

[07-Sep-2011 14:05:44] <Gat0rvean> gotcha

[07-Sep-2011 14:12:30] <jb> 2>&1> still isn't redirecting.

[07-Sep-2011 14:14:38] <rmatte> jb: show the command

[07-Sep-2011 14:15:26] <Gat0rvean> rmatte: does zenperfsnmp still run on the devices even after you remove the template?

[07-Sep-2011 14:15:46] <dhopp> Gat0rvean:  it might until the next config cycle

[07-Sep-2011 14:16:08] <dhopp> Gat0rvean:  you can restart zenpersnmp

[07-Sep-2011 14:16:11] <Jane_Curry> hiya

[07-Sep-2011 14:16:29] <dhopp> Hi Jane_Curry

[07-Sep-2011 14:16:42] <Gat0rvean> dhopp: I watched it for one full cycle, and it appears to be running on the same amount of devices

[07-Sep-2011 14:16:54] <Jane_Curry> I see you have been very industrious answering forum questions - good on you!

[07-Sep-2011 14:17:26] <Jane_Curry> Got a techie one....

[07-Sep-2011 14:17:33] <Hackman238> Jane_Curry: Been trying

[07-Sep-2011 14:17:42] <jb> rmatte: I'm just trying to redirect the output of zenperfsnmp ie, zenperfsnmp run -v10 -d device > file

[07-Sep-2011 14:17:45] <dhopp> Gat0rvean:  the daemons only check for new configuration every so often. 20 or 30 minutes by default I think

[07-Sep-2011 14:17:55] <Gat0rvean> dhopp: gotcha, thanks

[07-Sep-2011 14:18:13] <Jane_Curry> I want to create a Command that elicits input from the user using a read statement and then uses that as a variable in the script that is run

[07-Sep-2011 14:18:18] <jb> Hackman238: you guys doing anything with the XML stat channel on BIND?

[07-Sep-2011 14:18:19] <Jane_Curry> Anyone been there?????

[07-Sep-2011 14:18:59] <dhopp> Gat0rvean:  actually it looks like the default Config Cycle is 360 minutes.  You can also use the "Push Changes" in the gear menu if necessary

[07-Sep-2011 14:19:22] <dhopp> Jane_Curry:  I saw your forum post and I have no idea :-)

[07-Sep-2011 14:20:36] <Jane_Curry> <dhopp> Ho-hum - deep joy!

[07-Sep-2011 14:32:43] <Hackman238> jb: I dont follow

[07-Sep-2011 14:33:04] <Hackman238> Jane_Curry: Hum.

[07-Sep-2011 14:33:17] <Hackman238> Jane_Curry: You'd need to add a custom dialog for that

[07-Sep-2011 14:34:41] <jb> Hackman238: i guess.. are you monitoring BIND9?

[07-Sep-2011 14:35:28] <Jane_Curry> Hackman238: I was trying to avoid ZenPack country and just use standard stuff

[07-Sep-2011 14:35:50] <jb> hah, my ghetto parsing of XML using WGET is just that.. ghetto.

[07-Sep-2011 14:35:55] <Hackman238> jb: Oh, no I dont here. We do at OFT.

[07-Sep-2011 14:36:16] <jb> ah.

[07-Sep-2011 14:36:30] <Hackman238> jb: PM'd you a contact for extensive BIND monitoring

[07-Sep-2011 14:36:56] <Hackman238> Jane_Curry: I'm not certain how it could be done without a whole extra dialog.

[07-Sep-2011 14:37:21] <Jane_Curry> I can make it work by having my command as xterm -display :0.0 -e '/pathname/pathname1 ${TALES1} ${TALES2} '

[07-Sep-2011 14:37:36] <Jane_Curry> but my customer doesn't have xterm available

[07-Sep-2011 14:38:59] <Hackman238> Jane_Curry: So a xterm pops up that accepts user input?

[07-Sep-2011 14:43:59] <jb> hrm ok, now to figure out why this graph is not populating.

[07-Sep-2011 14:44:02] <jb> datasources tests fine

[07-Sep-2011 14:44:21] <Hackman238> jb: Check the rrd file to see if its got data

[07-Sep-2011 14:44:30] <Hackman238> jb: rrdtool last file.rrd

[07-Sep-2011 14:46:20] <jb> hm, no rrd.

[07-Sep-2011 14:46:24] <jb> but there is an image.

[07-Sep-2011 14:46:41] <jb> does that even make sense?

[07-Sep-2011 14:49:46] <Hackman238> jb: No, if the graph renders at all there must be an rrd

[07-Sep-2011 14:50:11] <rmatte> Hackman238: incorrect, the image can appear with a message stating that the RRD file doesn't exist

[07-Sep-2011 14:50:35] <Jane_Curry> hackman38:yup - I get an xterm that prompts for input and then uses it - all fine - except for organisations where there is no xterm!

[07-Sep-2011 14:50:59] <Hackman238> rmatte: I cant say I've ever seen that.

[07-Sep-2011 14:51:08] <Hackman238> jb: Turns out it can happen

[07-Sep-2011 14:51:13] <rmatte> I've seen it many times, it generally only occurs if the graph is the first graph

[07-Sep-2011 14:51:19] <rmatte> if it's not the first graph it doesn't happen

[07-Sep-2011 14:51:26] <Hackman238> rmatte: Ah gotcha

[07-Sep-2011 14:51:41] <Hackman238> jb: Does the graph have a comment in it suggesting the file doesnt exist?

[07-Sep-2011 14:51:53] <Hackman238> Jane_Curry: Windows env?

[07-Sep-2011 14:51:59] <rmatte> basically instead of showing "nan" values you get a "Missing RRD" message in it's place

[07-Sep-2011 14:52:18] <Hackman238> rmatte: Oh right, right. I've seen that. I stand corrected.

[07-Sep-2011 14:52:28] <rmatte>

[07-Sep-2011 14:52:48] <rmatte> I would have been surprised if you hadn't seen it before lol

[07-Sep-2011 14:54:12] <Hackman238> rmatte: Yeah IDK what I was thinking

[07-Sep-2011 14:54:22] <rmatte> hehe

[07-Sep-2011 14:54:42] <Hackman238> rmatte: I think its bad brain index- sort of associate "missing rrd" graph with no graph

[07-Sep-2011 14:55:06] <rmatte> could be

[07-Sep-2011 15:04:54] <jb> well, there i no message saying the file does not exist

[07-Sep-2011 15:05:06] <rmatte> but there is a rendered image there?

[07-Sep-2011 15:05:08] <jb> and I just searched $ZENHOME/perf/Devices/* for *ssl* , and its not there.

[07-Sep-2011 15:05:16] <jb> no

[07-Sep-2011 15:05:32] <rmatte> how do you know the filename will contain "ssl"?

[07-Sep-2011 15:05:47] <jb> well, thats the name of the datasource.

[07-Sep-2011 15:06:23] <rmatte> you are taking case sensitivity in to account, yeh?

[07-Sep-2011 15:06:35] <jb> yep

[07-Sep-2011 15:07:01] <Jane_Curry> Hackman238 Nope - Linux - RedHat

[07-Sep-2011 15:20:29] <jb> yeah, all other RRD's are named by datasource.

[07-Sep-2011 15:20:39] <jb> so I don't know why this one is mis-named, missing, etc.

[07-Sep-2011 15:25:24] <illsci> question..

[07-Sep-2011 15:25:37] <illsci> any advice on installing zennoss in ec2 as far as device discovery?

[07-Sep-2011 15:25:59] <willwh> illsci: I run zenoss from ec2

[07-Sep-2011 15:26:07] <willwh> you mean discovering ec2 devices?

[07-Sep-2011 15:26:12] <illsci> how do you do device discovery

[07-Sep-2011 15:26:14] <illsci> yeah

[07-Sep-2011 15:26:17] <illsci> from in zenoss...

[07-Sep-2011 15:26:19] <willwh> you can't

[07-Sep-2011 15:26:22] <illsci> I'm JUST setting it up now...

[07-Sep-2011 15:26:39] <illsci> So you manually add in ip's ?

[07-Sep-2011 15:26:50] <willwh> yes

[07-Sep-2011 15:27:05] <illsci> I think at the very least I could write some code... im not sure if zenoss has hooks for that

[07-Sep-2011 15:27:08] <willwh> snmp not allowed for ec2 too - so you can monitor it easily

[07-Sep-2011 15:27:17] <willwh> there is a zenpack if you have the ec2 monitoring stuff

[07-Sep-2011 15:27:26] <illsci> yeah I do..

[07-Sep-2011 15:27:28] <illsci> hmm

[07-Sep-2011 15:28:00] <willwh> docs/DOC-4423

[07-Sep-2011 15:28:24] <illsci> so you can't do snmp in ec2 or you can

[07-Sep-2011 15:28:35] <illsci> i see..

[07-Sep-2011 15:39:40] <Jane_Curry> <illsci> get the AWS ZenPack - that discovers your cloud machines automatically

[07-Sep-2011 15:40:33] <Jane_Curry> The ZenPack collector diacovers instances and then also  discovers nodes by public dns name and then links up the two - works quite well

[07-Sep-2011 15:42:37] <illsci> willwh: that zenpack doesn't seem to install enPacks.zenoss.ZenAWS-1.0.3-py2.6.egg.zip

[07-Sep-2011 15:43:22] <Jane_Curry> What version of Zenoss do you have - I have the 1.0.3-py2.6.egg.zip installed on Zenoss 3.1

[07-Sep-2011 15:43:39] <illsci> 3.2.0.0

[07-Sep-2011 15:43:48] <illsci> 3.2.0-0

[07-Sep-2011 15:43:52] <illsci> OSError: [Errno 20] Not a directory: '/usr/local/zenoss/zenoss/Products/ZenPacks.zenoss.ZenAWS-1.0.3-py2.6.egg/skins'

[07-Sep-2011 15:44:07] <illsci> thats the error I got trying to do zenpack --install filename.zip

[07-Sep-2011 15:44:10] <Jane_Curry> Ah - the new stuff - haven't gone there yet - how does it barf?

[07-Sep-2011 15:44:52] <Jane_Curry> You did unzip the ZenPack first????

[07-Sep-2011 15:45:13] <illsci> nope

[07-Sep-2011 15:45:42] <Jane_Curry> Unzip and then try and reinstall - will probably work then

[07-Sep-2011 15:45:46] <illsci> k

[07-Sep-2011 15:47:35] <jb> i think its the COMMAND datasource causing problems, but I can't get zenperfsnmp's output into a file so I can more closely look at it.

[07-Sep-2011 15:49:16] <illsci> it worked

[07-Sep-2011 15:49:22] <illsci> thanks... first zenpack I've ever installed

[07-Sep-2011 15:49:32] <jb> is there switch on zenperfsnmp to redirect output a file?

[07-Sep-2011 15:49:55] <Jane_Curry> Good - signing off now - dinner time here

[07-Sep-2011 15:50:35] <Hackman238> Jane_Curry: Later jane

[07-Sep-2011 15:51:05] <Hackman238> jb: zenperfsnmp run now will dump to the screen. Then redirect stdout and stderr as usual

[07-Sep-2011 15:51:05] <dhopp> jb: try stopping zenperfsnmp

[07-Sep-2011 15:52:23] <dhopp> jb: then zenperfsnmp run —cycle > file.log 2>&1

[07-Sep-2011 15:54:23] <jb> heh, ok, im stupid   thanks.

[07-Sep-2011 15:55:15] <psyklone> jb - pass in -v10 for debug level output, and if you have a specific device you know you want to test against, pass that device ID in with -d

[07-Sep-2011 15:55:40] <psyklone> zenperfsnmp run -v10 -d DEVICEID 2>&1 | tee /tmp/zenperfsnmp.out

[07-Sep-2011 15:57:03] <dhopp> jb: oops I forgot the -v10 :-)

[07-Sep-2011 15:57:09] <jb> right

[07-Sep-2011 15:57:27] <dhopp> but psyklone's way is probably a little better...

[07-Sep-2011 15:57:53] <Hackman238> I must be the only guy who still uses the old "run now" instead of -cycle LOL

[07-Sep-2011 15:57:57] <jb> gr, this doesn't make any sense.  I defined a new COMMAND datasource, and a graph.  the graph image is appearing, no date.  but, there is no .rrd for the datasource, and no mention in zenperfsnmp -v10.

[07-Sep-2011 15:58:16] <Hackman238> jb: command datasources are collected by zencommand

[07-Sep-2011 15:58:19] <willwh> illsci: see yuo got it sorted, good

[07-Sep-2011 15:58:20] <psyklone> if it's a command data source ...

[07-Sep-2011 15:58:23] <jb> ah, right.

[07-Sep-2011 15:58:25] <jb> haha.

[07-Sep-2011 15:58:25] <willwh> illsci: sorry was out to lunch...

[07-Sep-2011 15:58:26] <psyklone> yeah, Hackman238 beat me to it

[07-Sep-2011 15:58:29] <jb> (i -so- knew that)

[07-Sep-2011 15:58:35] <Hackman238>

[07-Sep-2011 15:58:51] <dhopp> Hackman238:  I thought —cycle basically ran zenperfsnmp in the foreground and "now" just runs once and exits…maybe I'm wrong

[07-Sep-2011 15:59:13] <jb> hm, not a lot of output from zencommand run -v10.

[07-Sep-2011 16:00:02] <jb> im thinking the datasource is getting ignored.

[07-Sep-2011 16:00:09] <Hackman238> jb: Really? zencommand should ahve more output

[07-Sep-2011 16:00:11] <jb> probably because it's not enabled.

[07-Sep-2011 16:00:12] <dhopp> jb: is it "enabled" ?

[07-Sep-2011 16:00:14] <jb> -gasp-

[07-Sep-2011 16:00:16] <dhopp> lol

[07-Sep-2011 16:00:17] <Hackman238> jb: LOL

[07-Sep-2011 16:00:49] <jb> it's working now

[07-Sep-2011 16:01:08] <Hackman238> dhopp: maybe IDK. I only ever run for one iteration in foreground

[07-Sep-2011 16:01:14] <dhopp> jb: it's amazing what happens when you tell zenoss to do what you want it to do :-P

[07-Sep-2011 16:01:27] <Hackman238> LOL

[07-Sep-2011 16:06:15] <jb> crazy, huh

[07-Sep-2011 16:06:48] <jb> i think it's working.. I'll check again in a few zencommand itterations.

[07-Sep-2011 16:07:19] <illsci> willwh: yeah it installed correctly.. now to try and use it

[07-Sep-2011 16:07:27] <illsci> willwh:  why not just use snmp in ec2?

[07-Sep-2011 16:07:34] <illsci> does that not work out well or something?

[07-Sep-2011 16:22:58] <dhopp> why is it only 3:22?  WHY?

[07-Sep-2011 16:23:19] <Simon4> move to the UK, it's 21:30 here

[07-Sep-2011 16:23:31] <Hackman238> dhopp: tell me about it

[07-Sep-2011 16:23:31] <Simon4> give or take a few min, anyway

[07-Sep-2011 16:25:19] <dhopp> psyklone:  you there?

[07-Sep-2011 16:25:40] <psyklone> dhopp, more or less

[07-Sep-2011 16:25:58] <dhopp> psyklone: how's the patch coming for 3.2?  Any progress?

[07-Sep-2011 16:30:48] <psyklone> dhopp, i'll have something for you in a few minutes

[07-Sep-2011 16:30:57] <dhopp> psyklone:  sweet

[07-Sep-2011 16:32:10] <Hackman238> psyklone: This the 'fix everything everyhere green button' patch? (see http://www.youtube.com/watch?v=IupmBK8K3l4 for ref.)

[07-Sep-2011 16:32:24] <Hackman238> LOL

[07-Sep-2011 16:32:31] <psyklone> lol ... no, now this pales by comparison

[07-Sep-2011 16:32:57] <Hackman238> hahaha

[07-Sep-2011 16:35:50] <dhopp> Hackman238:  ok..I think my co workers think I'm looney…I started to laughing outloud…I couldn't stop thinking "Now that was easy"

[07-Sep-2011 16:35:59] <dhopp> *to laugh

[07-Sep-2011 16:36:16] <Hackman238> dhopp: hahaha

[07-Sep-2011 16:36:42] <Hackman238> dhopp: I think Staples should sue Zenoss for ripping off the 'Easy Button'...all they did was pain dump it red and bam! LOL

[07-Sep-2011 16:36:47] <Hackman238> *paint

[07-Sep-2011 16:36:59] <Hackman238> *green

[07-Sep-2011 16:36:59] <dhopp> green you mean :-P

[07-Sep-2011 16:37:22] <Hackman238> dhopp: I'm not crazy!

[07-Sep-2011 16:37:39] <dhopp> so says the crazy man

[07-Sep-2011 16:37:51] <Hackman238> dhopp: LOL

[07-Sep-2011 16:38:03] <Hackman238> dhopp: Over worked is a better way to put it

[07-Sep-2011 16:40:00] <jb> hah, the damn graph still isn't updating.

[07-Sep-2011 16:40:35] <jb> http://pastebin.com/v3kS2CF0 .. you can see zencommand getting a value now.

[07-Sep-2011 16:40:45] <jb> 2011-09-07 15:17:38,763 DEBUG zen.zencommand: Output: '37\n'

[07-Sep-2011 16:41:47] <dhopp> jb: I have to ask the obvious…did you add the datapoint to the graph?

[07-Sep-2011 16:41:54] <jb> yes

[07-Sep-2011 16:42:15] <Hackman238> jb: Output: '37\n'

[07-Sep-2011 16:42:17] <Hackman238> jb: set it so the output is 'OK| datapointname=37'

[07-Sep-2011 16:42:26] <Hackman238> jb: Then pick nagios parser. Cati one always give me a hassle

[07-Sep-2011 16:42:31] <jb> ok sec.

[07-Sep-2011 16:44:09] <Hackman238> jb: Also create the datapoint as 'someName' where the name is meanful in the output. Ex: OK| pingMs=37. In this case, call the datapoint pingMs

[07-Sep-2011 16:44:10] <psyklone> dhopp, rmatte, Hackman238, etc. the patch and instructions for the leading-dot OID issue in 3.2 is on the trac ticket: http://dev.zenoss.com/trac/ticket/7859

[07-Sep-2011 16:44:26] <dhopp> psyklone:  taking  a look

[07-Sep-2011 16:44:45] <Hackman238> psyklone: TY

[07-Sep-2011 16:45:39] <jb> haha there we go

[07-Sep-2011 16:45:43] <dhopp> psyklone:  is it possible to use zenpatch?

[07-Sep-2011 16:45:43] <jb> thanks hackman

[07-Sep-2011 16:45:46] <jb> 2011-09-07 15:45:29,677 DEBUG zen.zencommand: Storing sslTps = 23.0 into Devices/fc-f5-03.follett.com/sslTps_sslTps

[07-Sep-2011 16:45:49] <dhopp> psyklone:  out of curiosity

[07-Sep-2011 16:45:51] <jb> 2011-09-07 15:45:29,675 DEBUG zen.zencommand: Output: 'OK| sslTps=23\n'

[07-Sep-2011 16:46:41] <psyklone> dhopp - not for this one, currently.

[07-Sep-2011 16:47:29] <Hackman238> jb: There ya go. Working

[07-Sep-2011 16:50:04] <dhopp> hrm

[07-Sep-2011 16:50:11] <dhopp> [zenoss@chi-rh5-test-02 Products]$ patch -p0 < r47525.patch

[07-Sep-2011 16:50:11] <dhopp> patching file ZenHub/services/SnmpPerfConfig.py

[07-Sep-2011 16:50:11] <dhopp> patch unexpectedly ends in middle of line

[07-Sep-2011 16:50:11] <dhopp> patch unexpectedly ends in middle of line

[07-Sep-2011 16:50:46] <dhopp> however the file gets updated

[07-Sep-2011 16:52:36] <psyklone> oh, that's probably a trailing line at the end. not a problem, but i'll check it in a sec

[07-Sep-2011 16:52:41] <jb> it would be cool if you could specify multiple "sources" or OID's per datasource, and then perform calculations on them.

[07-Sep-2011 16:53:42] <dhopp> jb: there is a community zenpack that does that I believe.  I think it's rmatte's zenpack actually.  docs/DOC-10224

[07-Sep-2011 16:53:58] <jb> oh, interesting. i'll check it.

[07-Sep-2011 16:54:09] <jb> yep cool

[07-Sep-2011 16:54:11] <dhopp> jb: however it currently uses zencommand so if you try to have it do a lot of calculations it has a lot of overhead

[07-Sep-2011 16:54:21] <dhopp> jb: he's working on daemonizing it

[07-Sep-2011 16:54:26] <jb> im just calling an external script via zencommand to do it currently..

[07-Sep-2011 16:55:13] <Hackman238> jb: Thats whatt he pack does. It in dev for daemonization

[07-Sep-2011 16:55:25] <Hackman238> Whoops. Yeah, what dhopp said

[07-Sep-2011 16:56:44] <jb> these types of calculations are necessary for some bigip stuff :/

[07-Sep-2011 17:01:33] <Hackman238> jb: No worries, its coming

[07-Sep-2011 17:02:00] <Hackman238> jb: In the mean time, have a laugh ->http://www.happyplace.com/3695/unintentionally-sexual-church-signs

[07-Sep-2011 17:03:07] <dhopp> Hackman238 / psyklone : for the zencommand problem of it stopping after the first datapoint…was it said earlier that a possible workaround (other then reverting Auto.py) was to force the parser to Nagios?

[07-Sep-2011 17:03:25] <Hackman238> dhopp: Yep

[07-Sep-2011 17:03:29] <dhopp> I was doing like 25 things at once so I might have missed that

[07-Sep-2011 17:03:35] <dhopp> Hackman238:  hrm..I might have found another bug

[07-Sep-2011 17:03:58] <dhopp> Hackman238:  I set Parser to Nagios. Click save, reopen datasource and parser is no longer set

[07-Sep-2011 17:04:10] <psyklone> dhopp, yes ... for now. fix is in progress on that one

[07-Sep-2011 17:04:28] <dhopp> psyklone:  I can't seem to get it to save the parser

[07-Sep-2011 17:04:43] <Hackman238> dhopp: Odd

[07-Sep-2011 17:07:41] <dhopp> Hackman238:  same thing happens on 3.1

[07-Sep-2011 17:08:02] <dhopp> I'm trying this on the Apache monitor template

[07-Sep-2011 17:09:54] <jb> Hackman238: hah.

[07-Sep-2011 17:10:21] <psyklone> dhopp, fyi - there was a trailing line i left on that, i've updated the patch that's on the ticket but it's fine as-applied as well

[07-Sep-2011 17:13:20] <Hackman238> Have to run

[07-Sep-2011 17:13:23] <Hackman238> Later all

[07-Sep-2011 17:19:29] <dhopp> psyklone:  Oids are being updated now..yay!

[07-Sep-2011 17:20:08] <psyklone> indeed

[07-Sep-2011 17:23:36] <dhopp> psyklone:  if I change the parser to Nagios in the UI it should still be set that way when I go back into the datasource right?

[07-Sep-2011 17:23:47] <psyklone> yessir

[07-Sep-2011 17:24:08] <dhopp> it's not…tried both in 3.1 and 3.2 for the Apache template

[07-Sep-2011 17:24:37] <psyklone> oh, ok. that's going to be a UI thing, but the parser should be saved (functionaly)

[07-Sep-2011 17:25:38] <dhopp> if that's the case, then that isn't a workaround to the problem.  I changed it to Nagios, saved and ran zencommand run -v10 and it stops after the first datapoint.

[07-Sep-2011 17:26:31] <dhopp> I should get my own trac login with all the tickets I'm going to be opening :-P

[07-Sep-2011 17:28:46] <psyklone> 0_0

[07-Sep-2011 17:34:19] kocolosk is now known as kocolosk|away

[07-Sep-2011 21:30:05] <illsci> hey anyone around?

[07-Sep-2011 21:30:25] <illsci> I'm curious how I force zenoss to discover a host over snmp

[07-Sep-2011 21:30:42] <illsci> the localhost... I just installed snmpd and now I want it to use snmp to checkout itself

[07-Sep-2011 21:32:22] <illsci> When I go to Infrastructure it shows "Loading, Please wait..." and it never returns

[07-Sep-2011 21:32:28] <illsci> what would cause that?

[07-Sep-2011 21:36:33] <stanchan> what does the logs say?

[07-Sep-2011 21:37:59] <illsci> which log?  the Z2.log shows the requests coming in

[07-Sep-2011 21:38:36] <stanchan> only 1 device or all devices?

[07-Sep-2011 21:38:57] <illsci> only 1 device

[07-Sep-2011 21:39:04] <illsci> I just installed this box

[07-Sep-2011 21:39:18] <illsci> and went through the wizard and didn't add anything

[07-Sep-2011 21:39:56] <stanchan> hmm... everything is setup correctly on the host?

[07-Sep-2011 21:40:05] <illsci> I think so

[07-Sep-2011 21:40:08] <stanchan> community string / net-snmp

[07-Sep-2011 21:40:15] <illsci> that might not be setup right

[07-Sep-2011 21:40:27] <illsci> but I'm surprised that the Loading... Please wait never returns

[07-Sep-2011 21:40:56] <illsci> It's like a modal.. you can't click on any of the left menu Device Servers stuff

[07-Sep-2011 21:41:07] <stanchan> yah... I never got that message before

[07-Sep-2011 21:41:20] <stanchan> but I'm not expert on zenoss

[07-Sep-2011 21:42:08] <stanchan> nothing in any log?

[07-Sep-2011 21:42:41] <stanchan> zenprocess?

[07-Sep-2011 21:43:28] <stanchan> zenmodeler?

[07-Sep-2011 22:05:13] crazed_ is now known as craed

[07-Sep-2011 22:05:17] craed is now known as crazed

[07-Sep-2011 22:05:47] crazed is now known as Guest75642

[07-Sep-2011 22:06:23] Guest75642 is now known as crazed

[08-Sep-2011 00:00:01] [disconnected at Thu Sep  8 00:00:01 2011]

[08-Sep-2011 00:00:02] [connected at Thu Sep  8 00:00:02 2011]

[08-Sep-2011 00:00:17] [zenoss-logger (logger bot) has joined #zenoss]

[08-Sep-2011 03:32:57] <froztbyte> :x...been typing up too much documentation

[08-Sep-2011 03:33:12] <froztbyte> keep autocorrecting 'decide' to 'device' when typing

[08-Sep-2011 05:28:44] <whyzgeek> any updates on ZCA new portal?

[08-Sep-2011 08:13:18] klone is now known as psyklone

[08-Sep-2011 08:20:58] <Hackman238> Morning all

[08-Sep-2011 08:21:25] <psyklone> howdy

[08-Sep-2011 08:24:09] <psyklone> 3.2 SNMP OID datasource issue - patch is linked in the trac ticket: http://dev.zenoss.com/trac/ticket/7859

[08-Sep-2011 08:24:35] <psyklone> 3.2 COMMAND data source issue - patch is linked in the trac ticket: http://dev.zenoss.com/trac/ticket/7860

[08-Sep-2011 08:25:20] <Hackman238> Very cool.

[08-Sep-2011 08:25:27] <Hackman238> psyklone: Hows by you?

[08-Sep-2011 08:25:33] <jmp242> Hi all from home - on occasionally

[08-Sep-2011 08:25:48] <Hackman238> jmp242: Hey, how goes it?

[08-Sep-2011 08:25:54] <psyklone> too early to tell, still getting the coffee flowing

[08-Sep-2011 08:25:58] <jmp242> well, Cornell is closed till 11 due to flooding

[08-Sep-2011 08:26:03] <Hackman238> jmp242: Oh wow.

[08-Sep-2011 08:26:04] <jmp242> so, not at work

[08-Sep-2011 08:26:08] <Hackman238> jmp242: Sorry to hear that

[08-Sep-2011 08:26:22] <Hackman238> jmp242: Your place flooded?

[08-Sep-2011 08:26:25] <jmp242> yea, didn't get up over the creek next to the lab

[08-Sep-2011 08:26:29] <jmp242> no, I'm up on a hill

[08-Sep-2011 08:26:41] <jmp242> but roads are closed so the university had to close

[08-Sep-2011 08:26:45] <Hackman238> jmp242: Glad you're still dry LOL

[08-Sep-2011 08:26:54] <Hackman238> jmp242: Gotcha.

[08-Sep-2011 08:27:12] <jmp242> yea - if we had flooded where I live, much of the eastern seaboard would be under water lol

[08-Sep-2011 08:27:24] <Hackman238> jmp242: Nick was telling me he's had soem flooding issues the last day or so too

[08-Sep-2011 08:27:31] <Hackman238> jmp242: LOL!

[08-Sep-2011 08:27:36] <jmp242> thogh my hill driveway does turn into a river

[08-Sep-2011 08:27:40] <Hackman238> psyklone: I hear that

[08-Sep-2011 08:28:25] <Hackman238> jmp242: Yeah I know how that is. My old place up in Salem, NY always ended up with a devistated driveway after heavy rains sice its a steep hill.

[08-Sep-2011 08:37:10] <Gat0rvean> is adding a new collector in Zenoss 3.X.X as easy as just Advanced > Collectors > Add Monitor?

[08-Sep-2011 08:37:38] <hmp> yes and no

[08-Sep-2011 08:37:54] <Gat0rvean> hmp: would you expound a bit on the no part?

[08-Sep-2011 08:38:06] <hmp> you have to prepare the collector with a user, default installation, ssh keys and environment

[08-Sep-2011 08:38:14] <hmp> also, install the required zenpack on the hub

[08-Sep-2011 08:38:40] <Gat0rvean> hmp: do you have a link to some documentation I could see for that specific job?

[08-Sep-2011 08:40:07] <hmp> docs/DOC-5861

[08-Sep-2011 08:40:17] <hmp> also, read rmatte's comments in the comments section

[08-Sep-2011 08:40:23] <jb> hrm, so I'm thinking about doing a fresh 4.0.2 install.

[08-Sep-2011 08:40:31] <jb> and not even trying to upgrade my 3.x instance.

[08-Sep-2011 08:40:37] <jb> if I can find some spare hardware.

[08-Sep-2011 08:40:46] <hmp> jb: is 4.x avalaible yet? like unstable or something? git?

[08-Sep-2011 08:40:53] <jb> no, its enterprise.

[08-Sep-2011 08:40:57] <jb> and it's out.

[08-Sep-2011 08:41:00] <hmp> nice

[08-Sep-2011 08:41:12] <jb> its called Service Dynamics/Resource Manager.

[08-Sep-2011 08:42:04] <jb> i just feel like there will be too much bs lingering from an upgrade.

[08-Sep-2011 08:42:09] <jb> like there was from 2->3

[08-Sep-2011 08:44:53] <jb> i feel like it was a silent release, though.

[08-Sep-2011 08:44:58] <jb> nobody told me about it.. no account manager, etc.

[08-Sep-2011 08:45:11] <Simon4> yeah, we had to go hunting for it

[08-Sep-2011 08:45:14] <Simon4> (also ent customer)

[08-Sep-2011 08:45:23] <jb> yeah, whats up with that?

[08-Sep-2011 08:45:48] <jb> Simon4: have you played with it yet?

[08-Sep-2011 08:46:06] <Simon4> jb: not the final version, but I played with one of the betas some

[08-Sep-2011 08:46:12] <Simon4> it's far better resource-wise

[08-Sep-2011 08:46:40] <jb> cool, any plans to upgrade?

[08-Sep-2011 08:47:26] <Simon4> I believe it's in the pipeline here - I'm out of the monitoring team now and doing other things, but I keep an ear out

[08-Sep-2011 08:47:33] <Simon4> we're still on 2.5 atm

[08-Sep-2011 08:47:36] <jb> ah

[08-Sep-2011 08:53:32] <Hackman238> jb: Why so?

[08-Sep-2011 08:55:10] <whyzgeek> upgrade from 3 to 4 is seemless but there are major changes to the architecture like adding rabbitmq and memcached and Java services plus relstorage

[08-Sep-2011 09:05:53] <Hackman238> jb: We're in the process of moving to 4.0.2 now. Generally speaking one should be able to poll 100-120% more datapoints per collector

[08-Sep-2011 09:06:10] <rocket> go for 200%

[08-Sep-2011 09:06:11] <rocket>

[08-Sep-2011 09:06:13] <Hackman238> jb: UI updates and the event system rework also allows for a useful event console and event history

[08-Sep-2011 09:06:24] <Hackman238> rocket: LOL. How goes it?

[08-Sep-2011 09:06:51] <rocket> it goes ..

[08-Sep-2011 09:07:12] <Hackman238> rocket: Glad to see you're in good spirits!

[08-Sep-2011 09:07:33] <rocket> lol if these are good spirits wait until my mood changes ..

[08-Sep-2011 09:07:49] <Hackman238> rocket: LOL

[08-Sep-2011 09:08:52] [disconnected at Thu Sep  8 09:08:52 2011]

[08-Sep-2011 12:00:02] [connected at Thu Sep  8 12:00:02 2011]

[08-Sep-2011 12:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[08-Sep-2011 12:02:49] <dhopp> Can somebody make my clients not request stupid things?  I'd appreciate it

[08-Sep-2011 12:14:20] <rmatte> ^ I second that

[08-Sep-2011 12:14:55] <rmatte> We're just finishing up this huge project for a big client, but they are nitpicking every detail

[08-Sep-2011 12:14:58] <rmatte> driving me nuts

[08-Sep-2011 12:17:06] <rmatte> I'll be so glad when it's over it's just a time drain

[08-Sep-2011 13:22:05] <Hackman238> rmatte: I hate it when that happens

[08-Sep-2011 13:22:31] <Hackman238> rmatte: Need to point to the SOW and say 'its not in the SOW, the next SOW we'll adress that problem'

[08-Sep-2011 13:41:09] <rmatte> the problem is that the people who write our SOWs are idiots and overpromise

[08-Sep-2011 13:41:20] <rmatte> which is something we're working on addressing

[08-Sep-2011 13:41:22] <Hackman238> rmatte: ...Zenoss PS?

[08-Sep-2011 13:41:31] <Hackman238> rmatte: LOL sorry, sorry, bad joke.

[08-Sep-2011 13:41:33] <rmatte> lol

[08-Sep-2011 13:42:43] kocolosk|away is now known as kocolosk

[08-Sep-2011 13:44:26] <tvincent_> Thats sales people for you. =)

[08-Sep-2011 13:44:41] <tvincent_> they dont' believe in under promise over deliver

[08-Sep-2011 13:45:43] <rmatte> yeh

[08-Sep-2011 13:46:06] <rmatte> "Capture the sun and deliver it to your doorstep?  Sure we can do that, would you like us to also grab the moon while we're up there?"

[08-Sep-2011 13:46:28] <rmatte> With the development team waving them off in the background

[08-Sep-2011 13:46:35] <tvincent_> actually I wanted a zenpack for monitoring solar flares that could simultaneously interface with Hubble.

[08-Sep-2011 13:46:44] <rmatte> lol

[08-Sep-2011 13:47:18] <rmatte> "Well you won't need that once you have the actual sun in your possession."

[08-Sep-2011 13:47:24] <rmatte> "It should only take a day or two."

[08-Sep-2011 13:49:50] <wayne___> any thoughts on a quick way to add a bunch of products (with product keys) under the Manufacturers hierarchy?  I'm trying to add them through zendmd, but I'm lost as to which method I need to call

[08-Sep-2011 13:49:55] <wayne___> methods, rather

[08-Sep-2011 13:54:00] <rocket> Hackman238: grrr another bad zenoss joke!

[08-Sep-2011 13:55:40] <Hackman238> rocket: doh! caught.

[08-Sep-2011 13:55:55] <Hackman238> rocket: This ones true though- but its PS/sales fault.

[08-Sep-2011 13:56:13] <Hackman238> rocket: Far as Im concerned support gets an NPS score of 9

[08-Sep-2011 13:56:17] <Hackman238> (10 being best)

[08-Sep-2011 13:56:38] <rocket> grrr it should be 11

[08-Sep-2011 13:56:49] otakup0p1 is now known as otakup0pe

[08-Sep-2011 13:56:49] <rocket> we work after hours when we are not supposed to

[08-Sep-2011 13:57:18] <Hackman238> rocket: I say 9 because groups who get a 10 often suffer regression because they think they've mastered everything.

[08-Sep-2011 13:57:38] <Hackman238> cluther: Hey chet.

[08-Sep-2011 14:03:58] <Hackman238> rocket: If I were your middle mgmt I'd mass eliminate uneeded levels of managers, high on extra developers and with the money saved give everyone raises.

[08-Sep-2011 14:04:33] <Hackman238> rocket: Sounds like mgmt will string you along up until you're just fed up with them.

[08-Sep-2011 14:05:50] <rocket> Hackman238: sounds like two wrong assumptions .. we are too small to have that many managers yet .. and actually I am quite happy with my boss etc ...

[08-Sep-2011 14:06:04] <Hackman238> rocket: Oh ok.

[08-Sep-2011 14:06:23] <rocket> we are no where near as large as a rackspace or dell etc where there are that many extra levels etc

[08-Sep-2011 14:06:30] <stanchan> any big bugs with 3.2 core that people are running into?  planning to do some testing on it and want to see if I should wait.

[08-Sep-2011 14:06:38] <Hackman238> rocket: I've heard the opposite from others, so I did just assume you were in the same catagory

[08-Sep-2011 14:06:42] <rocket> 2 levels and I am at CEO level .. lol

[08-Sep-2011 14:07:02] <Hackman238> stanchan: the topic line has links to patches for datasource problems

[08-Sep-2011 14:07:11] <Hackman238> rocket: Ah, yeah special situation for you then

[08-Sep-2011 14:07:14] <Hackman238> LOL

[08-Sep-2011 14:07:34] <jmp242> stanchan: I'd leave 3.2 myself for a 3.2.1

[08-Sep-2011 14:07:36] <jmp242> hopefully

[08-Sep-2011 14:07:44] <jmp242> but there are patches for 3.2 that makes it not broken

[08-Sep-2011 14:08:32] <Hackman238> rocket: Any who, again, dont think I mean anything directed at you  &  support.

[08-Sep-2011 14:09:55] <Gat0rvean> jmp242: speaking of patches

[08-Sep-2011 14:10:07] <Gat0rvean> I just upgraded my sandbox server and it stopped graphing

[08-Sep-2011 14:10:45] <Gat0rvean> so I applied the patch for RRD tool and got this:

[08-Sep-2011 14:10:48] <Gat0rvean> patching file ZenHub/services/SnmpPerfConfig.py

[08-Sep-2011 14:10:48] <Gat0rvean> patch unexpectedly ends in middle of line

[08-Sep-2011 14:10:48] <Gat0rvean> Reversed (or previously applied) patch detected!  Assume -R? [n] y

[08-Sep-2011 14:10:48] <Gat0rvean> Hunk #1 succeeded at 32 with fuzz 1.

[08-Sep-2011 14:22:34] <dhopp> Gat0rvean:  I'm taking a look.  I applied the patch yesterday

[08-Sep-2011 14:23:01] <Gat0rvean> dhopp: it seems to have worked, just a bit fuzzy on the output

[08-Sep-2011 14:28:21] <dhopp> Gat0rvean:  yeah the patch works…but it looks like it's missing a blank line at the end so you get that "unexpectedly ends in middle of line" message

[08-Sep-2011 14:30:27] <Gat0rvean> dhopp: gotcha, is there any reason to edit the file and re-apply the patch?

[08-Sep-2011 14:31:02] <dhopp> no

[08-Sep-2011 14:31:18] <Gat0rvean> dhopp: k, thanks

[08-Sep-2011 14:31:26] <stanchan> jmp242: thanks... since I build from source, I have no issues patching, but I'll hold off until 3.2.1 to push it to production

[08-Sep-2011 14:31:51] <rocket> stanchan: its very possible there will not be a 3.2.1

[08-Sep-2011 14:32:08] <rocket> stanchan: most likely we will be recommending these two patches with 3.2

[08-Sep-2011 14:32:24] <rocket> I have not heard anything official in this regard

[08-Sep-2011 14:32:45] <stanchan> rocket: works for me... thanks

[08-Sep-2011 14:32:51] <rocket> from a support perspective we are assisting our customers to install these two patches

[08-Sep-2011 14:33:07] <rocket> as we know who has downloaded this version.

[08-Sep-2011 14:41:44] <jmp242> oh ok

[08-Sep-2011 14:41:48] <jmp242> or the patches are fine too

[08-Sep-2011 14:42:07] <jmp242> hey, I'm not an RRD expert, but if you want to have (or try to have) non averaged data

[08-Sep-2011 14:42:17] <jmp242> your RRA needs to specify guage, and LAST?

[08-Sep-2011 14:42:29] <jmp242> This thread is where I'm trying to help

[08-Sep-2011 14:42:30] <jmp242> vmessage/61241#61241

[08-Sep-2011 14:42:34] <Thinh> http://i.imgur.com/2tBis.jpg

[08-Sep-2011 14:42:37] <Thinh> ws

[08-Sep-2011 14:44:06] <Hackman238> jmp242: LAST yes, but its still a rate

[08-Sep-2011 14:44:27] <Hackman238> jmp242: so they are unlikely to get whole numbers if thats what they are after

[08-Sep-2011 14:44:28] <jmp242> I thought as much, i.e. you're never going to prevent RRD averaging in some way

[08-Sep-2011 14:44:35] <Hackman238> jmp242: Exactly

[08-Sep-2011 14:48:08] <Hackman238> jmp242: Ah, Mr.Zwickel from OFT.

[08-Sep-2011 14:48:13] <Hackman238> (SteveZ)

[08-Sep-2011 14:48:14] <dhopp> rocket:  I'm going to be honest..I think that's stupid…telling new customers "to get zenoss to do what it's supposed to do you have to install 3.2 and then go through this convoluted patching process" is a horrible image.  at least make it so you can run zenpatch to do it

[08-Sep-2011 14:49:30] <dhopp> rocket: and on top of that saying "don't worry about the error messages patch throws…we just have a bad patch file but it patches ok"

[08-Sep-2011 14:50:16] <rmatte> dhopp: yeh, I basically just said the exact same thing in a ZCA email that I sent out to Nick

[08-Sep-2011 14:51:18] <rocket> I have no control over releases

[08-Sep-2011 14:51:41] <dhopp> rocket:  I'm not blaming you..but it's frustrating to see this

[08-Sep-2011 14:54:10] <rocket> dhopp: no offense taken, but do realize its very easy to make comments until you live the dream everyday from a support perspective

[08-Sep-2011 14:54:46] <jb> hrm, think I found some hardware for 4.0.2, hopefully.

[08-Sep-2011 14:55:41] <rmatte> Here's some text from the ZCA email: http://fpaste.org/xStB/raw/

[08-Sep-2011 14:55:44] <Hackman238> jb: Hardware?

[08-Sep-2011 14:55:52] <rmatte> for those who didn't receive it

[08-Sep-2011 14:59:06] <rmatte> I think that Zenoss needs to adopt more of a Blizzard style of "When it's done" development.  Quality over quantity.

[08-Sep-2011 14:59:26] <jb> Hackman238: yeah, im going to spin up a new 4.0.2 instance, instead of upgrading my 3.x instance.

[08-Sep-2011 14:59:40] <jb> it will give me a chance to walk through each device/platform and make sure everything is the way it should be.

[08-Sep-2011 14:59:42] <Hackman238> jb: Ah gotcha

[08-Sep-2011 14:59:53] <Hackman238> jb: Good practice

[08-Sep-2011 15:00:20] <rocket> rmatte: except then there are enterprise customers who are wanting new features all the time ... its a fine balancing act

[08-Sep-2011 15:00:41] <rocket> we are not a game company etc

[08-Sep-2011 15:00:45] <rmatte> rocket: I'm sure that demand could be met with ZenPacks that eventually get rolled in to releases

[08-Sep-2011 15:01:01] <dhopp> rmatte:  I'm ok with a quicker release cycle for Core but I'm interested in what QA was done.  These bugs were basically found within hours of release.

[08-Sep-2011 15:01:32] <rmatte> dhopp: the release was really rushed, they went from beta to release in 1 week, I've never seen them do that before

[08-Sep-2011 15:01:37] <rmatte> so I'd imagine not much testing was done

[08-Sep-2011 15:01:43] <jb> this enterprise customer has been asking for features for years.. and none of them have been added :/

[08-Sep-2011 15:02:09] <rmatte> jb: bribe the sales team to request them and you'll get them

[08-Sep-2011 15:02:11] <rmatte> lol

[08-Sep-2011 15:02:16] <rocket> jb: We try ..

[08-Sep-2011 15:02:17] <dhopp> rocket:  kind of what rmatte was getting at is the 'beta' cycle for this release was like 2 days..and it was a small handful of users that even knew about it.  I would imagine your enterprise 4.0 beta was months

[08-Sep-2011 15:02:29] <dhopp> rmatte:  I didn't mean from beta to release being fast...

[08-Sep-2011 15:02:32] <jb> my account manager didn't even let me know about 4.x and the entire re-branding of the product

[08-Sep-2011 15:02:40] <rocket> dhopp: I have no idea why the rushed 3.2 release .. I wasnt part of it

[08-Sep-2011 15:03:10] <rocket> jb: really? that doesnt make sense ..

[08-Sep-2011 15:03:25] <jb> i know.

[08-Sep-2011 15:03:26] <dhopp> rocket:  again..I don't mean to poke at you or offend you..but unfortunately you are the one that can hear me voice my frustration.

[08-Sep-2011 15:03:38] <jb> i found out in this IRC channel, and had to email rachel just to ask what the deal was..

[08-Sep-2011 15:04:43] <Thinh> huhu

[08-Sep-2011 15:04:56] <Thinh> its all about the $$$$

[08-Sep-2011 15:05:11] <JohnnyNOC> yea

[08-Sep-2011 15:05:18] <JohnnyNOC> i like #

[08-Sep-2011 15:05:18] <JohnnyNOC> $

[08-Sep-2011 15:05:19] <jb> well, its not like we are getting enterprise for free

[08-Sep-2011 15:05:28] <JohnnyNOC> i imagine everyone at zenoss with diamond watches

[08-Sep-2011 15:05:32] <JohnnyNOC> teeth, etc.

[08-Sep-2011 15:05:39] <Thinh> its all about more $$$$

[08-Sep-2011 15:06:05] <rocket> JohnnyNOC: oh nice .. I wish

[08-Sep-2011 15:06:06] <Thinh> i would be happy with zenoss if they kept 2.5 and keep fixing all the damn bugs =0

[08-Sep-2011 15:06:26] <JohnnyNOC> rocket it's fun to imagine

[08-Sep-2011 15:06:40] <rocket> you know many of the bugs are related to interacting with other tools etc

[08-Sep-2011 15:06:57] <rocket> eg specifications that were not followed on the other end

[08-Sep-2011 15:07:16] <rocket> broken windows apis, snmp specifications not followed

[08-Sep-2011 15:07:37] <rocket> + all the moving pieces to support all the different monitoring everyone wants

[08-Sep-2011 15:08:04] <rocket> its not a simple problem space .. I have never found it easy in all my previous years of sysadmin work prior to this job

[08-Sep-2011 15:12:36] <rmatte> I was just thinking today that if they had just given the 2.5 UI a facelift with some new graphics that it would be basically better than the new UI

[08-Sep-2011 15:13:01] <dhopp> rmatte:  I'll be honest..I like the new interface

[08-Sep-2011 15:13:20] <rmatte> the new UI has major problems with it, I mean, you can't copy multiple templates at once, has to be done one at a time.  You can't move device classes.  You have to go in to ZMI to do that

[08-Sep-2011 15:13:52] <rmatte> Overall the design of the new UI is nice, but it has several glaringly obvious limitations that weren't present in the old UI

[08-Sep-2011 15:14:04] <dhopp> rmatte:  could you move device classes in 2.x?  It's been a while since I've used it

[08-Sep-2011 15:14:11] <rmatte> dhopp: yeh, you can

[08-Sep-2011 15:14:34] <rmatte> "Move Organizers"

[08-Sep-2011 15:15:32] <rmatte> and you can move multiple classes at once just like templates

[08-Sep-2011 15:15:32] <rocket> rmatte I am on a call but would like to address some of these when I get off of the call

[08-Sep-2011 15:15:49] <rmatte> rocket: ok, I'll see if I can remember any others while I'm waiting, thanks

[08-Sep-2011 15:16:01] * rmatte fires up his trusty 3.x box

[08-Sep-2011 15:17:55] <jb> im not a fan of the new UI personally

[08-Sep-2011 15:18:13] <jb> theres just too many things that don't make sense or work properly.. iimo.

[08-Sep-2011 15:18:59] <rmatte> lol, I just remembered a bug that's still present in the new UI

[08-Sep-2011 15:19:22] <jb> one thing - you should be able to assign group/system/location directly from the add device dialog.

[08-Sep-2011 15:19:25] <rmatte> if you navigate to Advanced -> ZenPacks, then click on Daemons, you no longer see that you're in the Advanced section at the top

[08-Sep-2011 15:19:38] <rmatte> just a cosmetic bug

[08-Sep-2011 15:19:44] <jb> half of the time the windows services ajax-ish stuff flakes out

[08-Sep-2011 15:19:59] <jb> the event console.. well, we all know   i hear this is fixed in 4 though.

[08-Sep-2011 15:20:21] <rmatte> jb: you can

[08-Sep-2011 15:20:28] <rmatte> you need to click the More... link in the add device window

[08-Sep-2011 15:20:54] <rmatte> then you get the options for location, groups, systems, serial numbers, etc...

[08-Sep-2011 15:20:57] <jb> you sure?

[08-Sep-2011 15:21:03] <rmatte> I'm staring at it right now

[08-Sep-2011 15:21:05] <rmatte> want a screenshot?

[08-Sep-2011 15:21:06] <jb> im looking at it right now

[08-Sep-2011 15:21:06] <rmatte> lol

[08-Sep-2011 15:21:14] <jb> i don't see a box for location/system/group

[08-Sep-2011 15:21:23] <rmatte> there's a link you click

[08-Sep-2011 15:21:28] <rmatte> that says "More..."

[08-Sep-2011 15:21:32] <jb> yes

[08-Sep-2011 15:21:34] <jb> sec

[08-Sep-2011 15:21:39] <rmatte> just above the Add button

[08-Sep-2011 15:21:45] <rmatte> once you click that you should see those boxes

[08-Sep-2011 15:21:57] <rocket> rmatte: are you looking at a 3.2 system btw?

[08-Sep-2011 15:22:01] <rmatte> yes, I am

[08-Sep-2011 15:22:02] <jb> im not

[08-Sep-2011 15:22:13] <rmatte> so I guess that's something fixed in 3.2

[08-Sep-2011 15:22:18] <rocket> still on the call ..

[08-Sep-2011 15:22:36] <rmatte> though I could have swore it was there in 3.1

[08-Sep-2011 15:22:39] <jb> sec

[08-Sep-2011 15:22:48] <rmatte> k

[08-Sep-2011 15:23:03] <rmatte> keep in mind that I'm also looking at a core system, not enterprise

[08-Sep-2011 15:24:46] <jb> http://thunder.jbdesign.net/~jbaird/adddevice.png

[08-Sep-2011 15:24:53] <jb> err

[08-Sep-2011 15:25:09] <rmatte> The requested URL /~jbaird/adddevice.png was not found on this server.

[08-Sep-2011 15:25:10] <jb> http://thunder.jbdesign.net/~jbaird/zenoss/adddevice.png

[08-Sep-2011 15:25:11] <rmatte> lol

[08-Sep-2011 15:25:25] <rmatte> ok, that must have been fixed in 3.2 then

[08-Sep-2011 15:26:24] <jb> windows services hardly ever load correctly for me

[08-Sep-2011 15:26:34] <jb> sorting them is terribly slow

[08-Sep-2011 15:26:42] <jb> i hate dealing with windows services in 3.x

[08-Sep-2011 15:26:57] <jb> i'm hopefuly that all of these issues are fixed in 4

[08-Sep-2011 15:26:59] <rmatte> agreed, that's another big issues

[08-Sep-2011 15:27:01] <rmatte> issue*

[08-Sep-2011 15:27:27] <dhopp> jb: what problems do you have with windows services?

[08-Sep-2011 15:27:36] <rmatte> jb: http://oi54.tinypic.com/2lwqj3d.jpg

[08-Sep-2011 15:27:39] <jb> the ajax bs never loads correctly

[08-Sep-2011 15:27:47] <jb> and the fact that you see ALL windows services on each device now

[08-Sep-2011 15:27:53] <jb> instead of only the ones that you are currently monitoring

[08-Sep-2011 15:27:57] <rocket> have you cleared the browser cache?

[08-Sep-2011 15:27:58] <rmatte> yeh, not just the ones being monitored

[08-Sep-2011 15:28:10] <jb> rmatte: yep, thats what I'm talking about.

[08-Sep-2011 15:28:21] <rmatte> I really miss the option to hide items that aren't monitored

[08-Sep-2011 15:30:33] <rmatte> I can't really think of anything that bugs me quite as much as the copying limitations though

[08-Sep-2011 15:31:00] <rmatte> copying/moving

[08-Sep-2011 15:32:48] <jb> i wish you could copy/paste directly from the event console

[08-Sep-2011 15:33:33] <rmatte> actually, I'm going to really miss being able to copy/paste from the device list since I do it all the time in 2.5

[08-Sep-2011 15:33:41] <rmatte> guess I have to export each time I want to do it now

[08-Sep-2011 16:09:16] <JohnnyNOC> can anyone help me understand why a device with a critical alert (ip service smtp is down) doesn't show up in the dashboard?

[08-Sep-2011 16:09:37] <JohnnyNOC> it has normal priority like all of our other devices (Except printers) and i kind of expect to see it in my 'device issues' portlet on the dashboard

[08-Sep-2011 16:09:40] <JohnnyNOC> am i missing something?

[08-Sep-2011 16:09:52] <wayne___> I don't think it's considered a device issue - it's a service issue

[08-Sep-2011 16:11:00] <rmatte> JohnnyNOC: the device is in Production state?

[08-Sep-2011 16:11:23] <rmatte> also, was the device in production state when the first event came in?

[08-Sep-2011 16:20:59] <rocket> ok I am back

[08-Sep-2011 16:21:11] <rmatte> hi

[08-Sep-2011 16:22:17] <rmatte> ok, so two things that were possible in 2.5 interface that are no longer: moving classes (multiple at once) and moving/copying templates (multiple at once, now can only copy 1 at a time and can't move)

[08-Sep-2011 16:23:06] <rmatte> same goes for moving any other kind of organizer such as location, group, system

[08-Sep-2011 16:25:18] <rmatte> It can be done via ZMI obviously, but it's ridiculous having to do it that way

[08-Sep-2011 16:26:13] <rmatte> afk for a sec, grabbing some chips

[08-Sep-2011 16:26:29] <illsci> So I just setup Zenoss and I'm trying to get it to monitor itself with snmp.  This host is in the ec2 cloud.  I setup snmp and can do manual snmpwalks but zenoss isn't working out yet..   When I click on Infrastructure it hangs forever showing "Loading Please wait"

[08-Sep-2011 16:26:38] <illsci> I saw this in the daemon logs through the web interface

[08-Sep-2011 16:26:47] <illsci> http://dpaste.com/611438/

[08-Sep-2011 16:27:02] <illsci> I'm not sure how to fix this yet...

[08-Sep-2011 16:28:39] <illsci> mnt/zenoss/common/bin/snmpwalk -v 1 -c public localhost system shows me a bunch of info...

[08-Sep-2011 16:28:50] <illsci> so I think snmpd is working...

[08-Sep-2011 16:30:37] <rmatte> illsci: you installed Zenoss 3.2?

[08-Sep-2011 16:30:51] <illsci> yeah

[08-Sep-2011 16:31:11] <rmatte> Use 3.1 instead, 3.2 has major bugs in it that should have been caught prior to release but weren't

[08-Sep-2011 16:31:24] <illsci> werd...

[08-Sep-2011 16:31:28] <rmatte> there should be patches for 3.2 soon if there aren't already (I haven't heard of any)

[08-Sep-2011 16:31:41] <illsci> can you get a ubuntu zenoss-stack deb for 3.1?

[08-Sep-2011 16:31:52] <rmatte> oh wait, looks like there might be patches

[08-Sep-2011 16:31:53] <rmatte> hold on

[08-Sep-2011 16:32:10] <psyklone> those patches should be zenpatchable shortly, btw

[08-Sep-2011 16:32:24] <rmatte> oh, so they aren't yet?

[08-Sep-2011 16:32:47] <rmatte> illsci: yes, you can get any package type for 3.1

[08-Sep-2011 16:33:05] <psyklone> they are patchable via diffs, but not available as automatically patch 'zenpatch' patches yet.

[08-Sep-2011 16:33:06] <illsci> hmm I wonder how I can specify an exact version with apt

[08-Sep-2011 16:33:10] <hmp> here are some patches mentioned earlier today:

[08-Sep-2011 16:33:11] <hmp> 14:24 < psyklone> 3.2 SNMP OID datasource issue - patch is linked in the trac ticket: http://dev.zenoss.com/trac/ticket/7859                                                                     ben6

[08-Sep-2011 16:33:15] <hmp> 14:24 < psyklone> 3.2 COMMAND data source issue - patch is linked in the trac ticket: http://dev.zenoss.com/trac/ticket/7860

[08-Sep-2011 16:33:16] <rmatte> illsci: don't use apt... hold on

[08-Sep-2011 16:34:08] <rmatte> illsci: 32 or 64bit linux?

[08-Sep-2011 16:34:12] <illsci> 32

[08-Sep-2011 16:34:22] <illsci> m1.small instance

[08-Sep-2011 16:34:50] <rmatte> http://sourceforge.net/projects/zenoss/files/zenoss-3.1/zenoss-3.1.0/zenoss-stack_3.1.0_i386.deb/download

[08-Sep-2011 16:35:09] <rmatte> or you can just wait for the patches to become zenpatchable for 3.2

[08-Sep-2011 16:35:16] <rmatte> but you can always upgrade from 3.1 to 3.2 whenever it is

[08-Sep-2011 16:35:38] <rmatte> hmp: trying to walk someone through patching via a .patch file is a pain

[08-Sep-2011 16:36:40] <rocket> rmatte: I am sorry I have had two urgent calls come in that I am juggling

[08-Sep-2011 16:36:46] <rmatte> I personally hope they decide to do a 3.2.1 release fix the bugfixes in place or we're going to be running in to this issue a lot

[08-Sep-2011 16:36:59] <rmatte> rocket: all good, I understand

[08-Sep-2011 16:37:43] <rmatte> with the bugfixes in place rather

[08-Sep-2011 16:39:06] <rmatte> illsci: If you install 3.1 make sure you uninstall 3.2 first

[08-Sep-2011 16:39:23] <rmatte> or you could run in to all sorts of fun issues

[08-Sep-2011 16:49:53] <Sashness> how the hell do i view the configuration properties for a device in the zope database?

[08-Sep-2011 16:50:10] <Sashness> i want to write  a report that gets ssCpuRawWait but apparently its not named that in the actuall zope database

[08-Sep-2011 16:50:27] <rmatte> Sashness: is that a datapoint?

[08-Sep-2011 16:50:53] <rmatte> the value isn't stored in Zope if it is, it's stored in an RRD file and you need to use the getRRDValue function to retrieve it

[08-Sep-2011 16:50:56] <rmatte> i.e:

[08-Sep-2011 16:51:06] <rmatte> d = dmd.Devices.findDevice('whatever')

[08-Sep-2011 16:51:16] <rmatte> d.getRRDValue('ssCpuRawWait')

[08-Sep-2011 16:52:20] <rmatte> which will retrieve the most current value

[08-Sep-2011 16:52:31] <rmatte> if you need to retrieve averages over a period of time it gets a bit more complicated

[08-Sep-2011 16:53:07] <rmatte> These are scripts I wrote that dump similar data and accept input of start/end dates as well as epoch time values to do it: docs/DOC-4664

[08-Sep-2011 16:53:12] <rmatte> you can use them as reference

[08-Sep-2011 16:54:45] <rmatte> if you're actually scripting a report, you can use existing reports as examples

[08-Sep-2011 16:55:00] <rmatte> they've made the scripting pretty simple in more recent versions of Zenoss (2.5+)

[08-Sep-2011 16:56:48] <Sashness> yeah i copied the CPU utilization and tried to modify it to remove load average and add in CPU Wait but i don't know the value to add there

[08-Sep-2011 16:56:58] <Sashness> as ssCpuRawWait doesn't seem to work

[08-Sep-2011 16:57:35] <rmatte> can you show me your code?

[08-Sep-2011 16:57:38] <rmatte> fpaste.org

[08-Sep-2011 17:03:12] <Hackman238> Later all!

[08-Sep-2011 17:03:17] <rmatte> later

[08-Sep-2011 17:09:42] <Sashness> http://pastebin.com/6vYuMAJM

[08-Sep-2011 17:15:09] <rmatte> k I think you're missing the point, there's also a plugin file that you have to edit which is what actually gathers the data

[08-Sep-2011 17:15:26] <rmatte> objects python:here.ReportServer.plugin('cpu', tableState);

[08-Sep-2011 17:15:30] <rmatte> it'll be called cpu.py

[08-Sep-2011 17:15:35] <rmatte> and it'll be in a plugins folder

[08-Sep-2011 17:15:47] <rmatte> plugins directory rather, should be under ZenReports

[08-Sep-2011 17:16:12] <rmatte> you'll also need to restart Zenoss to apply the changes to the report

[08-Sep-2011 17:16:33] <rmatte> and if you're adding a new report you need to do: python $ZENHOME/Products/ZenReports/ReportLoader.py --force

[08-Sep-2011 17:16:35] <rmatte> as the zenoss suer

[08-Sep-2011 17:16:37] <rmatte> user*

[08-Sep-2011 17:17:15] <rmatte> well sorry, you'll have to restart zope to apply the changes at the very least

[08-Sep-2011 17:17:17] <rmatte> zopectl restart

[08-Sep-2011 17:17:24] <rmatte> afk for a few, washroom

[08-Sep-2011 17:17:52] <dhopp> rmatte:  good to know

[08-Sep-2011 17:17:56] <dhopp> :-)

[08-Sep-2011 17:29:55] <rmatte> back

[08-Sep-2011 17:30:08] <rmatte>

[08-Sep-2011 17:30:40] <Thinh> dat azz

[08-Sep-2011 17:31:03] <rmatte> :|

[08-Sep-2011 17:56:46] <illsci> clear

[08-Sep-2011 17:56:51] <Sam-I-Am> fail

[08-Sep-2011 17:56:57] <rmatte> hehe

[08-Sep-2011 17:57:02] <illsci> hehe

[08-Sep-2011 17:57:03] <nexex_> ^L

[08-Sep-2011 17:57:04] <Sam-I-Am> sup rmatte

[08-Sep-2011 17:57:04] <rmatte> missing a slash there

[08-Sep-2011 17:57:17] <rmatte> Sam-I-Am: not much, was just about to leave work

[08-Sep-2011 17:57:39] <Sam-I-Am> sounds like a good plan

[08-Sep-2011 17:58:04] <rmatte> let's not kid ourselves, sounds like a *great* plan

[08-Sep-2011 17:58:06] <rmatte>

[08-Sep-2011 17:58:08] <Sam-I-Am> i'm in telco hell today... sprint phone started acting up last night and i can prove its a cell tower, not my phone... but sprint still wants to replace my phone and... yeah.

[08-Sep-2011 17:58:18] <rmatte> lame

[08-Sep-2011 17:58:24] <rmatte> I just ordered a new phone yesterday

[08-Sep-2011 17:58:24] <Sam-I-Am> its not under warranty, so $$$ and it wont fix it

[08-Sep-2011 17:58:29] <rmatte> going to be switching carriers finally

[08-Sep-2011 17:58:32] <illsci> Is there anyone here from zenoss?

[08-Sep-2011 17:58:47] <illsci> I'm curious why the apt repo doesn't allow you to install the past versions... http://dev.zenoss.org/deb/dists/main/stable/binary-i386/

[08-Sep-2011 17:58:50] <rmatte> which means I'll have unlimited talk, text, data, and long distance to canada/US for $40/month

[08-Sep-2011 17:59:03] <rmatte> which is way better than the crap plan that I'm on for $75/month right now

[08-Sep-2011 17:59:23] <illsci> do you get unlimited data right now for that 75?

[08-Sep-2011 17:59:40] <rmatte> illsci: I have no idea, I never use the apt repo, just a tip, the standard stack installer (the .bin), all the .deb is is that wrapped up in a .deb

[08-Sep-2011 17:59:51] <rmatte> I always personally install using the standard stack installer

[08-Sep-2011 18:00:02] <rmatte> you don't want automatic updates to Zenoss, trust me

[08-Sep-2011 18:00:09] <illsci> I understand

[08-Sep-2011 18:00:15] <rmatte> you want to test and decide when you upgrade, or you'll be bashing your head on a table

[08-Sep-2011 18:00:16] <rmatte> lol

[08-Sep-2011 18:00:19] <illsci> thats why I want to be able to specify the version

[08-Sep-2011 18:00:25] <illsci> with apt

[08-Sep-2011 18:00:38] <rmatte> you shouldn't even use apt

[08-Sep-2011 18:00:40] <illsci> and in the future I'll use puppet to enforce the specific version

[08-Sep-2011 18:00:45] <illsci> rmatte... why not

[08-Sep-2011 18:00:54] <illsci> why not use the package management system that is used for the os im using

[08-Sep-2011 18:00:56] <rmatte> well, are you deploying to more than one system?

[08-Sep-2011 18:01:00] <illsci> when zenoss makes packages for you

[08-Sep-2011 18:01:28] <rmatte> the packages aren't really what you'd expect from a deb though, it's literally all dependencies rolled in to one, so there's no point in using apt

[08-Sep-2011 18:01:30] <illsci> I'm just setting it up manually the first time to see it work and then im goign to write some puppet manifests

[08-Sep-2011 18:02:01] <rmatte> if you're using puppet to deploy then it makes sense, other than that it doesn't

[08-Sep-2011 18:02:18] <rmatte> I use Ubuntu Server for my Zenoss boxes too, but I don't bother with aptitude since it doesn't do anything useful for you

[08-Sep-2011 18:02:26] <rmatte> it's not like it's managing external dependencies or anything

[08-Sep-2011 18:02:30] <rmatte> it's literally just one package

[08-Sep-2011 18:02:38] <illsci> i get that

[08-Sep-2011 18:02:59] <illsci> my only problem is that the zenoss repo doesn't seem to allow you to install older releases...

[08-Sep-2011 18:03:02] <rmatte> when I want to upgrade I just wget the new .bin, chmod +x it, and execute it

[08-Sep-2011 18:03:06] <illsci> i think its because of the way the repo is managed

[08-Sep-2011 18:03:08] <rmatte> it takes care of the upgrade, and I'm done

[08-Sep-2011 18:03:22] <illsci> I dont do manual system administration

[08-Sep-2011 18:03:34] <rmatte> you'd have to talk to Nick Yeates (nyeates) about problems with the repo

[08-Sep-2011 18:03:35] <illsci> I try to never have to log into a box

[08-Sep-2011 18:03:37] <rmatte> he's the community manager

[08-Sep-2011 18:03:40] <illsci> werd

[08-Sep-2011 18:03:45] <stanchan> I assume you have a different position for RPMS ;-)

[08-Sep-2011 18:03:59] <illsci> I try to never use redhat

[08-Sep-2011 18:04:05] <rmatte> RPMs do actually manage the external dependencies somewhat (MySQL), so they make sense

[08-Sep-2011 18:04:18] <rmatte> but I'm not a big fan of redhat personally

[08-Sep-2011 18:04:27] <illsci> my last job was a big redhat shop

[08-Sep-2011 18:04:31] <Sam-I-Am> ahhhh crikey, irssi just hung

[08-Sep-2011 18:04:38] <Sam-I-Am> i think electronics hate me today

[08-Sep-2011 18:04:39] <illsci> but I would just blow servers away and rebuild them

[08-Sep-2011 18:04:42] <stanchan> hehe... I'm mantain 1000 custom RPMs, so I am used to it

[08-Sep-2011 18:04:54] <rmatte> I maintain 2 custom debs

[08-Sep-2011 18:04:56] <illsci> stanchan: there was guys at my old job that did that

[08-Sep-2011 18:04:58] <rmatte> lol

[08-Sep-2011 18:05:06] <willwh> Sam-I-Am: irssi hung? wtf you running it on?

[08-Sep-2011 18:05:06] <illsci> I used to hate how they did that stuff...

[08-Sep-2011 18:05:23] <rmatte> willwh: unisys 286

[08-Sep-2011 18:05:26] <Sam-I-Am> willwh: the same box i've run it fine on for a loooong time

[08-Sep-2011 18:05:30] <illsci> sometimes you have to to have certain options that arent in the os distros packages.. but I would say maintain them in the open and bug the people who make them to add them

[08-Sep-2011 18:05:31] <illsci> so you dont have to

[08-Sep-2011 18:05:40] <stanchan> The only way to manage that is with a build system like koji... it is a pain without it

[08-Sep-2011 18:05:45] <willwh> yah I run mine on my debian box under my floor :]

[08-Sep-2011 18:05:49] <willwh> irssi and screen ftwwwwwwwww

[08-Sep-2011 18:06:03] <Sam-I-Am> thats what i'm using

[08-Sep-2011 18:06:04] <rmatte> bitchx + screen is my flavour

[08-Sep-2011 18:06:08] <willwh> =)

[08-Sep-2011 18:06:10] <illsci> mine too

[08-Sep-2011 18:06:11] <Sam-I-Am> i think screen might have been the culprit

[08-Sep-2011 18:06:14] <willwh> Sam-I-Am: do you use screen_away?

[08-Sep-2011 18:06:17] <Sam-I-Am> connectivity at this coffee shop is shitty at best

[08-Sep-2011 18:06:20] <willwh> ah

[08-Sep-2011 18:06:24] <rmatte> screen_away?

[08-Sep-2011 18:06:32] <willwh> screen_away.pl (it's an irssi script)

[08-Sep-2011 18:06:37] <rmatte> what does it do?

[08-Sep-2011 18:06:46] <Sam-I-Am> their router box terminates random sessions

[08-Sep-2011 18:06:47] <willwh> it marks you away when you detach your screen session

[08-Sep-2011 18:06:49] <willwh> it's sweet

[08-Sep-2011 18:07:02] <rmatte> ah, I don't like marking myself away/back, it's too spammy

[08-Sep-2011 18:07:07] <willwh> you can have it nick chnage too - I should edit the script to only nick change on specific servers

[08-Sep-2011 18:07:09] <rmatte> if people want to know whether or not I'm around they can ask me

[08-Sep-2011 18:07:12] <willwh> but I am not that great with perl ;]

[08-Sep-2011 18:07:12] * rmatte is away,  gone ( l! off)  : os/ bx

[08-Sep-2011 18:07:13] <Sam-I-Am> i just attach and detach, no one knows.

[08-Sep-2011 18:07:13] * rmatte is back,  rejoice ( g! 1s)  : os/ bx

[08-Sep-2011 18:07:17] <Sam-I-Am> lol bx

[08-Sep-2011 18:07:21] <Sam-I-Am> you haxx0r

[08-Sep-2011 18:07:23] <rmatte> bx + osiris script

[08-Sep-2011 18:07:29] <rmatte> it's what I'm used to

[08-Sep-2011 18:07:30] <willwh> rmatte: nono, it doesn't state anything

[08-Sep-2011 18:07:34] <willwh> i.e.

[08-Sep-2011 18:07:36] <willwh> whois me

[08-Sep-2011 18:07:38] <Sam-I-Am> ircII ftw

[08-Sep-2011 18:07:44] <willwh> then say you have, and I'll detach my session

[08-Sep-2011 18:07:50] <willwh> and you can whois me again ;>

[08-Sep-2011 18:07:58] <willwh> it does not print to channel though

[08-Sep-2011 18:08:21] <willwh> great script though :]

[08-Sep-2011 18:08:24] <rmatte> willwh: yeh, I understand the standard /away command

[08-Sep-2011 18:08:29] <willwh>

[08-Sep-2011 18:08:29] <rmatte> lol

[08-Sep-2011 18:08:44] <rmatte> I've been on IRC going on like 14 years now

[08-Sep-2011 18:08:49] <rmatte>

[08-Sep-2011 18:08:54] <willwh> me too

[08-Sep-2011 18:09:01] <rmatte> nice

[08-Sep-2011 18:09:05] <willwh> I love all the new kids (I op on a few servers for gaming stuff)

[08-Sep-2011 18:09:09] <willwh> they are like "

[08-Sep-2011 18:09:13] <Sam-I-Am> i tend not to /whois people

[08-Sep-2011 18:09:14] <willwh> what is this irc you speak of

[08-Sep-2011 18:09:21] * willwh facepalms

[08-Sep-2011 18:09:31] <Sam-I-Am> what is this irc you speak of?

[08-Sep-2011 18:09:36] * Sam-I-Am uses microsoft comic chat

[08-Sep-2011 18:09:38] <willwh> internet relay chat

[08-Sep-2011 18:09:39] <willwh>

[08-Sep-2011 18:09:46] <Sam-I-Am> anyone remember that thing?

[08-Sep-2011 18:09:46] <rmatte> haha, I miss microsoft comic chat

[08-Sep-2011 18:09:55] <Sam-I-Am> <user> 234957234@#%$@#%#@%#@$%#@crap  hello

[08-Sep-2011 18:10:04] <rmatte> "I'm a cartoon alien and I'm chatting!"

[08-Sep-2011 18:10:48] <rmatte> have to admit though, it was an entertaining idea at the time

[08-Sep-2011 18:11:06] <rmatte> but IRC should never die in my opinion

[08-Sep-2011 18:11:11] <rmatte> since there's really nothing better out there

[08-Sep-2011 18:11:11] <illsci> i agree..

[08-Sep-2011 18:11:19] <Sam-I-Am> we'll just keep getting older

[08-Sep-2011 18:11:21] <willwh> no there isn't

[08-Sep-2011 18:11:23] <willwh> yup

[08-Sep-2011 18:11:28] <illsci> i had a guy trying to tell me everyone has moved on to twitter

[08-Sep-2011 18:11:29] <Sam-I-Am> and bitching about the young folks

[08-Sep-2011 18:11:33] <willwh> and the average IRCer age will be 65.

[08-Sep-2011 18:11:38] <rmatte> illsci: haha

[08-Sep-2011 18:11:43] <willwh> yah

[08-Sep-2011 18:11:47] <willwh> "real men user twitter"

[08-Sep-2011 18:11:52] <Sam-I-Am> problem with most other chat venues is they a) dont handle >2 people well and b) arent persistent

[08-Sep-2011 18:11:54] <willwh> I have heard the same thing from ppl

[08-Sep-2011 18:12:00] <willwh> exactly.

[08-Sep-2011 18:12:11] <Sam-I-Am> and i cant screen them

[08-Sep-2011 18:12:12] <willwh> and.... let's face it, bots are awesome too

[08-Sep-2011 18:12:14] <willwh> hooray for eggdrop

[08-Sep-2011 18:12:18] <Sam-I-Am> lol eggdrop

[08-Sep-2011 18:12:23] <Sam-I-Am> that brings back memories of... 1995

[08-Sep-2011 18:12:29] <rmatte> yeh, I've setup quite a few eggdrops in my days

[08-Sep-2011 18:12:30] <willwh> TCL, TCL, woooo \o/

[08-Sep-2011 18:12:45] <rmatte> yeh, I still code a little in tcl from time to time

[08-Sep-2011 18:12:49] <Sam-I-Am> op wars, invasions of foreigners, channel takeovers

[08-Sep-2011 18:12:59] <willwh> I am actually using this now: http://drupal.org/project/bot for one site

[08-Sep-2011 18:13:09] <Sam-I-Am> lots of channel forking due to disagreements

[08-Sep-2011 18:13:22] <Sam-I-Am> i think the general community has evolved a bit

[08-Sep-2011 18:13:29] <Sam-I-Am> although i tend to hang out in stable channels

[08-Sep-2011 18:13:34] <rmatte> I'm quite sure I could write my own IRC bot from scratch, the protocol really isn't that difficult

[08-Sep-2011 18:13:41] <Sam-I-Am> PING PONG

[08-Sep-2011 18:13:52] * willwh brings his paddle

[08-Sep-2011 18:13:52] <Sam-I-Am> (done irc via telnet to 6667)

[08-Sep-2011 18:13:54] <rmatte> I've connected to IRC servers via telnet before and successfully chatted lol

[08-Sep-2011 18:14:16] <willwh> quassel was not bad too when I was testing it

[08-Sep-2011 18:14:22] <rmatte> the annoying part is when the server pings you, if you don't pong back on time it disconnects you

[08-Sep-2011 18:14:26] <rmatte> having to do it by hand is a pain

[08-Sep-2011 18:14:31] <Sam-I-Am> forums are good for persistence, but not live

[08-Sep-2011 18:14:37] <Sam-I-Am> yeah

[08-Sep-2011 18:14:39] <Sam-I-Am> with a random number

[08-Sep-2011 18:14:42] <rmatte> yeh

[08-Sep-2011 18:14:44] <willwh> indeed - and that's often the argument between community managers

[08-Sep-2011 18:14:46] <willwh> "we need forums"

[08-Sep-2011 18:14:48] <willwh> "we need irc"

[08-Sep-2011 18:14:55] <rmatte> they need both

[08-Sep-2011 18:14:55] <Sam-I-Am> irc is supposed to log

[08-Sep-2011 18:14:55] <rmatte> lol

[08-Sep-2011 18:15:06] <Sam-I-Am> although it needs a de-bullshitter filter

[08-Sep-2011 18:15:17] <rmatte> well, IRC bots are supposed to log

[08-Sep-2011 18:15:17] <Sam-I-Am> like... only log stuff rmatte says

[08-Sep-2011 18:15:20] <rmatte> or clients

[08-Sep-2011 18:15:35] <rmatte> you'd just see a lot of cursing

[08-Sep-2011 18:16:01] <rmatte> lol, anywho, I'm out, later folks

[08-Sep-2011 18:16:05] <willwh> cat irclog/whatever.log | grep rmatte> >. weeeeeeeee.log

[08-Sep-2011 18:16:07] <Sam-I-Am> mmmkay, time to go back home and deal with sprint

[08-Sep-2011 18:16:08] <willwh> cyas o/

[08-Sep-2011 18:16:11] <Sam-I-Am> see y'all later

[08-Sep-2011 19:45:00] <illsci> man I can't win... even 3.1.0 shows nothing under devices

[08-Sep-2011 19:45:12] <illsci> instead of always showing its loading its just blank

[08-Sep-2011 19:45:38] <illsci> I gotta be doing something stupid

[08-Sep-2011 20:06:36] kocolosk is now known as kocolosk|away

[08-Sep-2011 20:31:09] kocolosk|away is now known as kocolosk

[08-Sep-2011 22:03:03] kocolosk is now known as kocolosk|away

[08-Sep-2011 22:10:31] kocolosk|away is now known as kocolosk

[09-Sep-2011 00:00:01] [disconnected at Fri Sep  9 00:00:01 2011]

[09-Sep-2011 00:00:02] [connected at Fri Sep  9 00:00:02 2011]

[09-Sep-2011 00:00:14] [zenoss-logger (logger bot) has joined #zenoss]

[09-Sep-2011 01:45:29] kocolosk is now known as kocolosk|away

[09-Sep-2011 01:48:17] kocolosk|away is now known as kocolosk

[09-Sep-2011 01:57:01] kocolosk is now known as kocolosk|away

[09-Sep-2011 04:26:54] <konvert> anyone online who can help me with something?

[09-Sep-2011 07:34:16] <konvert> hi dhopp

[09-Sep-2011 07:44:53] <konvert> Can someone help me to install pyodbc?

[09-Sep-2011 07:46:25] <illsci> morning

[09-Sep-2011 07:52:31] <konvert> morning

[09-Sep-2011 07:54:55] <illsci> Do you know if the Zenoss infrastructure page will be blank if zensnmp isn't working?

[09-Sep-2011 07:55:18] <hmp> no

[09-Sep-2011 07:55:29] <hmp> (it won't)

[09-Sep-2011 07:55:34] <illsci> When I click on Infrastructure I don't even see any menu for Devices

[09-Sep-2011 07:56:19] <hmp> thats not zenperfsnmp related

[09-Sep-2011 07:56:31] <illsci> any ideas where to start looking

[09-Sep-2011 07:56:36] <illsci> I downgraded from 3.2.0 to 3.1

[09-Sep-2011 07:56:55] <hmp> i never had that kind of problem

[09-Sep-2011 07:57:02] <illsci> I can't even add more devices because the devices menu isn't even there

[09-Sep-2011 07:57:06] <hmp> you checked the logs?

[09-Sep-2011 07:57:09] <illsci> I might uninstall and reinstall and see if that makes any difference

[09-Sep-2011 07:57:21] <illsci> yeah...I just see zensnmp show that its failing one job

[09-Sep-2011 08:02:40] <illsci> when you add your first device what's the best thing to do...

[09-Sep-2011 08:02:53] <illsci> Should you add localhost... or some other external device

[09-Sep-2011 08:37:14] <konvert> illsci are u there?

[09-Sep-2011 08:42:09] <illsci> yeah

[09-Sep-2011 08:42:14] <illsci> yeah I don't get it...

[09-Sep-2011 08:42:21] <illsci> I just uninstalled and reinstalled and didn't add a device

[09-Sep-2011 08:42:39] <illsci> and my Infrastructure -> Device page shows nothing... not even a button or menu or anything to add devices

[09-Sep-2011 08:43:48] <konvert> i also did a restore to 3.1 also no infrastructure...

[09-Sep-2011 08:44:01] <illsci> So 3.2.0 has bugs

[09-Sep-2011 08:44:11] <illsci> and 3.1.0 must also have some issue or I'm doing something wrong

[09-Sep-2011 08:44:15] <illsci> what's everyone actually using

[09-Sep-2011 08:44:31] <illsci> nagios?

[09-Sep-2011 08:44:57] <illsci> I really want to use this puppet-zenoss module

[09-Sep-2011 08:45:01] <rocket> 3.2 has patches for the 2 main bugs

[09-Sep-2011 08:45:06] <psyklone> illsci, there are patches for the two 3.2 issues that were reported this weekend

[09-Sep-2011 08:45:17] <konvert> rocket which ones?

[09-Sep-2011 08:45:29] <illsci> hmm well let me go back to them then... do you know where the docs for the patches are or how they should be applied

[09-Sep-2011 08:45:47] <psyklone> one sec, they are now available via zenpatch. i'll get the #'s

[09-Sep-2011 08:46:01] <rocket> konvert: see the topic

[09-Sep-2011 08:46:39] <dpetzel> when will a new build including the patches be released?

[09-Sep-2011 08:47:17] <rocket> dpetzel: I dont know

[09-Sep-2011 08:48:06] <psyklone> illsci, instructions for zenpatching the two issues mentioned above in 3.2:

[09-Sep-2011 08:48:07] <psyklone> 1. log in to the Zenoss server as the 'zenoss' user

[09-Sep-2011 08:48:07] <psyklone> 2. issue the following two commands to automatically download and apply these two patches:

[09-Sep-2011 08:48:07] <psyklone> 

[09-Sep-2011 08:48:07] <psyklone> zenpatch 26334

[09-Sep-2011 08:48:07] <psyklone> zenpatch 26335

[09-Sep-2011 08:48:08] <dpetzel> given the high level of complaints, I'd think getting those two patches included in the downloadable build would be a great thing to get done quicky

[09-Sep-2011 08:48:26] <illsci> sweet

[09-Sep-2011 08:48:54] <dpetzel> for folks like myself, where our zenoss boxes dont have inet access, IE zenpatch wont work, its a small PITA to have download the upgrade bits, upgrade, and then download/manually apply each patch

[09-Sep-2011 08:49:29] <rocket> dpetzel: grab the changeset put it in $ZENHOME/Products as r26334.patch for example

[09-Sep-2011 08:49:41] <rocket> then zenpatch as normal and it will bypass the download portion

[09-Sep-2011 08:50:16] <dpetzel> understood, its just an uncessary step and if folks are downloading the upgrade, and not following the thread, it creates frustrations.

[09-Sep-2011 08:50:37] <dpetzel> lucky for me I've been to busy to look at 3.2 this week, so I had the advantage of watching the threads pan out

[09-Sep-2011 08:50:51] <rocket> dpetzel: I am sorry but really this isnt helping. We understand the frustration and at this point its beating a dead horse

[09-Sep-2011 08:51:13] <dpetzel> sorry, I didnt mean to beat a dead horse at all, I was actually very impressed with the turnaround time on the patch

[09-Sep-2011 08:51:14] <rocket> dpetzel: we have provided work arounds

[09-Sep-2011 08:51:16] <illsci> rocket: do you guys know who manages the zenoss apt repo

[09-Sep-2011 08:51:25] <illsci> I was curious why they don't allow you to install previous versions

[09-Sep-2011 08:51:41] <dpetzel> was just adding my support ofr a small incremental download build. I have no complaints at all

[09-Sep-2011 08:51:43] <illsci> It would be nice to be able to do apt-get install zenoss-stack=3.1.0 or whatever

[09-Sep-2011 08:51:47] <dpetzel> sorry if it came across as negative

[09-Sep-2011 08:52:26] <rocket> dpetzel: I am sorry if i am defensive, this channel has become very negative lately and based on the work of my coworkers I find it a little frustrating

[09-Sep-2011 08:52:42] <rocket> there is so much work that happens on this stuff that many have no idea happens

[09-Sep-2011 08:53:05] <dpetzel> rocket: no worries, I no how it goes. Releases are rarely are "3 cheers" event, and its suck when no one recongizes all the GOOD that actually happened

[09-Sep-2011 08:53:15] <dpetzel> cause one or two populate bugs pop up

[09-Sep-2011 08:53:17] <rocket> 3.2 was meant to be an olive branch to the community to get a whole bunch of fixes from 3.1 out

[09-Sep-2011 08:53:48] <rocket> based on internal schedules 3.2 may have actually taken a lot longer to get out otherwise

[09-Sep-2011 08:54:26] <dpetzel> yeah I'm excited about many of the fixes included. had I not been so swamped this week I would have started upgrading

[09-Sep-2011 08:54:27] <rocket> yes there were two somewhat severe bugs that entered but I think we have done a good job of getting them fixed

[09-Sep-2011 08:54:55] <dpetzel> 100% agreed turn around and forum feedback has been amazing

[09-Sep-2011 08:55:13] <psyklone> dpetzel, that's true. and in this case, the bugs popped up very quickly and have patches available already ... which i think is good. 3.2 has a lot of fixes that are definitely worth having, and now that the two perf monitoring bugs are addressed there's no reason not to go to 3.2

[09-Sep-2011 08:55:17] <konvert> rocket or psyklone where can i find the patches and what they resolve

[09-Sep-2011 08:55:27] <rocket> the main reason you guys are still going to want 3.2 is an invalidation bug fix that should really improve performance of the backend

[09-Sep-2011 08:55:48] <rocket> konvert: see the topic

[09-Sep-2011 08:55:48] <psyklone> konvert - the patches are linked in the bugs referenced in the channel topic ... and are also available via zenpatch: 1. log in to the Zenoss server as the 'zenoss' user

[09-Sep-2011 08:55:48] <psyklone> 2. issue the following two commands to automatically download and apply these two patches:

[09-Sep-2011 08:55:48] <psyklone> 

[09-Sep-2011 08:55:48] <psyklone> zenpatch 26334

[09-Sep-2011 08:55:48] <psyklone> zenpatch 26335

[09-Sep-2011 08:57:08] <dpetzel> psyklone: agreed, I plan on running it through the paces next week. I'd love to have the file I download to include the two major patches, but I've been following the threads and channel close enough to know what to apply

[09-Sep-2011 09:00:26] <psyklone> dpetzel - there is talk of a 3.2.1 maintenance patch, however there's not one tentatively scheduled. i believe the thought at this point is to let 3.2 + patches run for a week or two to make sure nothing else big pops up that made it through QA, and then probably release a maintenance release. that avoids putting out a 3.2.1 right now that includes these two fixes, only to find another one or two bugs potentially that also need to be patched a few days la

[09-Sep-2011 09:00:26] <psyklone> ter

[09-Sep-2011 09:01:15] <dpetzel> psyklone: cool, its always tricky to know *when* to release maintenance patches. Its pretty much the wrong choice no matter when you pick LOL

[09-Sep-2011 09:01:23] <psyklone> indeed

[09-Sep-2011 09:06:18] <DW-10297> Is there a way to make Zenoss remove 'fixed' issues from the events list? I.e. if there is an issue and then it goes away then automatically clear it?

[09-Sep-2011 09:06:36] <jmp242> DW-10297: it already does that

[09-Sep-2011 09:06:45] <jmp242> assuming it can tell that the issue is fixed

[09-Sep-2011 09:08:09] <DW-10297> Ah, for example there was a report that said the smtp process was down on a host it said it started on 06/15 and ended on 06/15 but it was still displayed on the list of events

[09-Sep-2011 09:08:18] <DW-10297> with a big ole red x

[09-Sep-2011 09:09:08] <DW-10297> Also if I remove a host from infrastructure it still shows the events for that host.

[09-Sep-2011 09:10:44] <konvert> is it correct that the patching just takes a fraction of a second?

[09-Sep-2011 09:10:59] <konvert> do i need to restart something after patching?

[09-Sep-2011 09:12:24] <illsci> installing the zenoss-stack deb packages takes sooooooo long

[09-Sep-2011 09:12:36] <illsci> i wonder why... or if its because its an ec2 instance

[09-Sep-2011 09:14:11] <illsci> 3.2.0 installed and patched... starting

[09-Sep-2011 09:14:55] <jmp242> DW-10297: hmm

[09-Sep-2011 09:14:59] <rocket> is there a specific step taking so long?

[09-Sep-2011 09:15:01] <konvert> graphs still broken after clicking on it

[09-Sep-2011 09:15:13] <jmp242> That usually only happens if it doesn't get a clear event

[09-Sep-2011 09:15:20] <rocket> konvert: restart zenoss after patching

[09-Sep-2011 09:16:40] <konvert> rocket: done a 'service zenoss-stack restart' still graphs broken

[09-Sep-2011 09:17:13] <rocket> broken as in?

[09-Sep-2011 09:17:29] <konvert> red cross,

[09-Sep-2011 09:17:55] <konvert> and some graphs are working but after you click on them, also red cross

[09-Sep-2011 09:17:57] <rocket> are you seeing tracebacks on the collectors that would be generating the graph for the device?

[09-Sep-2011 09:18:15] <rocket> have you wiped the browser cache?

[09-Sep-2011 09:18:22] <rocket> check the zenrender logs for errors

[09-Sep-2011 09:18:37] <DW-10297> Also is it normal to see Zenoss hitting the same box's SNMPd several times a second?

[09-Sep-2011 09:18:57] <rocket> yes

[09-Sep-2011 09:19:01] <konvert> another issue : in multiprocessor servers, only 1 processor is monitored

[09-Sep-2011 09:19:06] <rocket> we make individual snmp calls

[09-Sep-2011 09:19:19] <rocket> because its more efficient than a walk for example

[09-Sep-2011 09:19:32] <rocket> so we usually make X calls to snmp per polling cycle

[09-Sep-2011 09:19:33] <DW-10297> right but i mean, constantly? I thought it only hit the box periodically

[09-Sep-2011 09:19:47] <rocket> X depends on the number of snmp datapoints you have

[09-Sep-2011 09:20:17] <rocket> it probably shouldnt be constant .. but I would look at zenperfsnmp logs in debug mode for clues

[09-Sep-2011 09:21:02] <konvert> rocket dont have a zenrender log

[09-Sep-2011 09:21:28] <rocket> do you have a zenrender daemon?

[09-Sep-2011 09:21:49] <rocket> perhaps I am forgetting and that is an enterprise only thing .. its been a while since I have looked at a core instance

[09-Sep-2011 09:22:00] <konvert> nope

[09-Sep-2011 09:22:11] <konvert> no zenrender daemon

[09-Sep-2011 09:24:07] <rocket> konvert: then I guess look at zenhub for tracebacks

[09-Sep-2011 09:25:06] <rocket> konvert: sorry I cant be of more help at the moment .. I have 9 other things to look at right now .. sorry

[09-Sep-2011 09:25:23] <konvert> rocket: and the strange thing is it only happens with my windows servers

[09-Sep-2011 09:25:40] <rocket> are you sure the rrd files even exist?

[09-Sep-2011 09:25:58] <rocket> eg are the windows daemons even creating the appropriate data point files?

[09-Sep-2011 09:26:11] <konvert> rocket how can i check?

[09-Sep-2011 09:26:41] <rocket> $ZENHOME/perf/Devices/<device> ... look for the rrds

[09-Sep-2011 09:27:54] <illsci> If you run mnt/zenoss/common/bin/snmpwalk -v 1 -c public localhost .1.3.6.1.4.1.2021.2 and it just shows END OF MIB

[09-Sep-2011 09:27:58] <konvert> yes i still have a folder os and then some .rrd files for the cpu's memory and sysuptime, 4cpu, 3meory, 1 uptime

[09-Sep-2011 09:27:59] <illsci> what's that mean...

[09-Sep-2011 09:28:12] <illsci> that snmp read only access isn't working?  or I don't have that mib installed

[09-Sep-2011 09:30:02] <konvert> rocket forget what i say about only the windows servers

[09-Sep-2011 09:30:40] <konvert> have it with some of the graphs of my linux servers

[09-Sep-2011 09:31:09] <rocket> sorry konvert like I said look for tracebacks in the logs as a first step.  I dont have any other time to look at this right now

[09-Sep-2011 09:31:37] kocolosk|away is now known as kocolosk

[09-Sep-2011 09:32:11] <konvert> the prolbem is i'm in europe, it is 15:30 now and my days end at 17:30 in US it's morning, and i dont get any response until i go home at evening...

[09-Sep-2011 09:32:54] <DW-10297> anyone know why Zenoss keeps generating alerts for hosts I removed awhile ago?

[09-Sep-2011 09:33:12] <rocket> konvert: sorry I have to work on enterprise cases right now .. I feel that pain but I really cant help you right now

[09-Sep-2011 09:34:16] <konvert> no problem, anyone else? otherwise i try it again on monday, and work on something other now

[09-Sep-2011 09:35:18] <konvert> DW-10297 because the hosts you removed, still sends traps to zenoss? try removing the zenoss machine in the hosts as trap receiver

[09-Sep-2011 09:35:48] <DW-10297> I don't use traps, I only use snmpd for zenoss to reach out to the hosts.

[09-Sep-2011 09:36:07] <DW-10297> it keeps reporting that the http process is down on a server I removed an hour ago

[09-Sep-2011 09:36:24] <DW-10297> when I click on the server in the alert it gives me an error obviously because I removed it =)

[09-Sep-2011 09:36:57] <JohnnyNOC> i have that same issue

[09-Sep-2011 09:37:19] <JohnnyNOC> super annoying - seems bouncing zenoss forces them to stop

[09-Sep-2011 09:37:39] <DW-10297> is there a way to configure it to bounce itself every time i remove something?

[09-Sep-2011 09:37:52] <JohnnyNOC> i suppose you could script that, but i don't know if it's a great idea

[09-Sep-2011 09:37:57] <DW-10297> (and yuck)

[09-Sep-2011 09:37:59] <JohnnyNOC> perhaps there's a real solution to that problem

[09-Sep-2011 09:38:00] <JohnnyNOC>

[09-Sep-2011 09:38:20] <JohnnyNOC> i mean, with time, they seem to go away as well..  but i have experienced exactly what you're describing

[09-Sep-2011 09:38:28] <JohnnyNOC> deleted devices are still being alerted on

[09-Sep-2011 09:38:55] <DW-10297> seems like they would want to fix something like that, but it is community software so whatever =)

[09-Sep-2011 09:39:52] <rocket> DW-10297: what version of zenoss are you using?

[09-Sep-2011 09:40:56] <DW-10297> i'll check

[09-Sep-2011 09:42:03] <SDuensin> FrIdAy!

[09-Sep-2011 09:42:06] <DW-10297> Zenoss 3.1.0

[09-Sep-2011 09:42:39] <whyzgeek> This is an old problem, you will need to restart the collector to fetch new configs

[09-Sep-2011 09:42:57] <DW-10297> is it fixed in 3.2?

[09-Sep-2011 09:43:18] <JohnnyNOC> how do you restart the collector?

[09-Sep-2011 09:43:21] <JohnnyNOC> you mean like, zenperfsnmp?

[09-Sep-2011 09:43:51] <whyzgeek> login to collector do zenoss restart

[09-Sep-2011 09:43:53] <whyzgeek> or update it

[09-Sep-2011 09:43:55] <rocket> yes ... it should be better in 3.2 but really isnt fixed until the collectors are on the new framework

[09-Sep-2011 09:44:03] <rocket> which happened in enterprise 4

[09-Sep-2011 09:44:24] <rocket> so it should be resolved when community 4 comes out

[09-Sep-2011 09:44:43] <whyzgeek> will there be any community 4?

[09-Sep-2011 09:45:15] <rocket> as far as I know its a work in progress now, nick knows more

[09-Sep-2011 09:45:22] <rocket> i dont deal with those day to day details

[09-Sep-2011 09:52:56] <whyzgeek> anybody knows an easy way to aggregate a number of OIDs and save them as separate value in RRDs? I would like to set threshold on the aggregate.

[09-Sep-2011 09:56:42] <photon01> Hi every1, my dashboard just crashed, any1 can help with that ?

[09-Sep-2011 09:56:57] <Sam-I-Am> morning guys

[09-Sep-2011 09:57:03] <Sam-I-Am> (and girls, if any)

[09-Sep-2011 09:57:06] <Sam-I-Am> mmm, zenoss girls

[09-Sep-2011 09:57:15] * Sam-I-Am fantasizes

[09-Sep-2011 09:57:48] <rocket> Sam-I-Am: as long as they are not like "Peggy" on the credit card commercial?

[09-Sep-2011 09:58:06] <Sam-I-Am> lol

[09-Sep-2011 09:58:10] * Sam-I-Am thinks of tesla girls

[09-Sep-2011 09:58:14] <Sam-I-Am> (80s reference)

[09-Sep-2011 09:58:24] <Sam-I-Am> guess that band

[09-Sep-2011 10:04:23] <jmp242> morning

[09-Sep-2011 10:13:43] <konvert> morning jmp242

[09-Sep-2011 10:13:57] <konvert> are u busy with something?

[09-Sep-2011 10:21:46] <jmp242> sadly yes

[09-Sep-2011 10:30:38] <kocolosk> i've noticed some funny behavior of zenprocess.  When the signature of a process I'm monitoring changes (e.g., because I changed some config flags and restarted it), zenprocess starts firing and clearing events on every monitoring cycle.  Restarting zenprocess caused it to settle down.

[09-Sep-2011 10:31:44] <kocolosk> The error event looks like "Unable to read processes on device foo; error: process-name 98a441bc859f8b2c8ba61488133a692d"

[09-Sep-2011 10:32:06] <kocolosk> and the clear event is just "Process table up for device foo"

[09-Sep-2011 10:32:20] <kocolosk> this is zenoss core 3.2

[09-Sep-2011 10:33:40] <kocolosk> anyone seen that behavior before?

[09-Sep-2011 11:02:58] <jmp242> kocolosk: yea you are probably not ignoring arguments ...

[09-Sep-2011 11:03:16] <jmp242> or it may be because of the way zenprocess seems to need to be restarted to pick up changes . . .

[09-Sep-2011 11:04:00] <kocolosk> jmp242: correct, i am not ignoring arguments, because I did want to match on one of the arguments in the pattern

[09-Sep-2011 11:04:27] <kocolosk> both the old process and the new process satisfied the regex

[09-Sep-2011 11:05:10] <kocolosk> it certainly does seem like zenprocess needs a restart to properly deal with changes like these

[09-Sep-2011 11:06:13] <kocolosk> though, curiously, I don't think the event flapping started happening till the first modeling of each device after the process in question restarted

[09-Sep-2011 11:09:32] <jmp242> Hmmm

[09-Sep-2011 11:09:39] <jmp242> not necessarily suprising either

[09-Sep-2011 11:09:46] <kocolosk> heh

[09-Sep-2011 11:10:34] <jmp242> zenoss doesn't notice configuration changes without a remodel

[09-Sep-2011 11:13:00] <kickehy_> is there any documentation/post upgrade tasks i need to know about after upgrading to 3.2?

[09-Sep-2011 11:13:07] kickehy_ is now known as kickehy

[09-Sep-2011 11:14:46] <kocolosk> jmp242: yeah, it's all a bit awkward.  At 0630 I see a "Process not running / Process up" event pair corresponding to the actual restart of the process.  The "Process up" event has the old command line arguments

[09-Sep-2011 11:15:30] <kocolosk> At 0830 the device is modeled and I start getting a persistent "Process not running" warning event with the *new* command line arguments (though the process is actually running)

[09-Sep-2011 11:15:48] <jmp242> interesting

[09-Sep-2011 11:16:01] <jmp242> There are issues with zenprocess

[09-Sep-2011 11:16:05] <kocolosk> I also start getting these "Unable to read processes / Process table up" event pairs every 180 seconds

[09-Sep-2011 11:16:24] <kocolosk> eventually i kick zenprocess and all is well

[09-Sep-2011 11:16:32] <jmp242> yea

[09-Sep-2011 11:16:50] <jmp242> I recall at some point there's a lot of zenprocess work done, but that may be for 4

[09-Sep-2011 11:17:04] <kocolosk> fair enough

[09-Sep-2011 11:17:41] <kocolosk> the workaround is pretty simple

[09-Sep-2011 11:18:01] <kocolosk> do you think it's the sort of thing that merits a trac ticket?

[09-Sep-2011 11:18:16] <kocolosk> not inclined to go to the trouble if it's a known/resolved thing

[09-Sep-2011 11:24:27] <jmp242> Well, I'd search trac

[09-Sep-2011 11:24:33] <jmp242> I expect there are tickets on things like this

[09-Sep-2011 11:25:08] <kocolosk> yeah, ok

[09-Sep-2011 11:26:52] <konvert> guys my graphs aren't broken anymore thx to you all

[09-Sep-2011 11:57:19] <photon01> how can I access my dashboard porlet list ?

[09-Sep-2011 11:57:33] <photon01> from zendmd or zope managemnt page

[09-Sep-2011 11:59:52] <JohnnyNOC> i was under the impression when a device was in the maintenance production state alerts wouldn't be generated...  apparently i'm wrong

[09-Sep-2011 11:59:56] <JohnnyNOC> can someone confirm?

[09-Sep-2011 12:00:01] [disconnected at Fri Sep  9 12:00:01 2011]

[09-Sep-2011 12:00:02] [connected at Fri Sep  9 12:00:02 2011]

[09-Sep-2011 12:00:14] [zenoss-logger (logger bot) has joined #zenoss]

[09-Sep-2011 12:00:19] <JohnnyNOC> i had some 'decommissioned' devices that were part of a maintenance window and when their production state changed to 'maintenance' alerts started kicking off

[09-Sep-2011 12:00:33] <JohnnyNOC> i put in a transform to suppress them but am wondering if there's osmething i'm missing or have misconfigured?

[09-Sep-2011 12:14:00] <dhopp> JohnnyNOC: I believe when they are in maintenance the events are still generated..and then for your alerting rules you can make sure the that state is not included so e-mails don't go out

[09-Sep-2011 12:15:07] <JohnnyNOC> *nod*

[09-Sep-2011 12:15:10] <JohnnyNOC> that's what i saw

[09-Sep-2011 12:15:20] <JohnnyNOC> i don't even want them hitting the dashboard which is why i setup the transform

[09-Sep-2011 12:15:31] <JohnnyNOC> which i don't believe is a terrible idea... comments/

[09-Sep-2011 12:15:38] <jmp242> maintenance does do monitoring

[09-Sep-2011 12:15:49] <jmp242> but it's totally for the alerting rules as far as I can tell

[09-Sep-2011 12:15:53] <jmp242> and some filtering in the GUI

[09-Sep-2011 12:15:59] <jmp242> doesn't affect event creation

[09-Sep-2011 12:18:19] <JohnnyNOC> *nod*

[09-Sep-2011 12:18:22] <JohnnyNOC> thanks guys

[09-Sep-2011 12:18:29] <JohnnyNOC> or girls

[09-Sep-2011 12:18:35] <JohnnyNOC>

[09-Sep-2011 12:43:23] <dhopp> can tell it's Friday..I have no desire to do anything I'm supposed to

[09-Sep-2011 12:48:31] <psyklone> dhopp --> +1

[09-Sep-2011 12:52:40] <JohnnyNOC> +1

[09-Sep-2011 13:00:29] <photon01> any1 can help me here ?

[09-Sep-2011 13:01:24] <dhopp> photon01:  just ask your question

[09-Sep-2011 13:01:57] <photon01> <dhopp> : I inserted some bad URL in a Site Window portlet and now I'm stuck with a blank Dashboard

[09-Sep-2011 13:02:28] <dhopp> photon01:  like completely blank?  You don't have any buttons / menus?

[09-Sep-2011 13:03:06] <photon01> <dhopp> : not like that. The usual dashboard page minus all portlets and the add portlet button

[09-Sep-2011 13:03:20] <photon01> <dhopp> : that would be a grey page..

[09-Sep-2011 13:03:37] <dhopp> photon01:  can you login as a different user?

[09-Sep-2011 13:03:47] <dhopp> photon01:  does the same thing happen?

[09-Sep-2011 13:04:20] <photon01> <dhopp> : no

[09-Sep-2011 13:04:33] <photon01> <dhopp> : it's ok when I change user

[09-Sep-2011 13:04:54] <photon01> <dhopp> : seems like all I need is to remove that portlet instance, I spent hours trying using zendmd

[09-Sep-2011 13:05:07] <photon01> <dhopp> : don't know the right method

[09-Sep-2011 13:06:10] <dhopp> photon1: I'm not sure how to do it..but you could probably delete the user and then recreate the user

[09-Sep-2011 13:06:25] <dhopp> photon01:  however..if you have a bunch of alerting rules you will lose them

[09-Sep-2011 13:06:45] <photon01> <dhopp> : actually... That's the admin user...

[09-Sep-2011 13:06:50] <dhopp> ack

[09-Sep-2011 13:06:54] <photon01>

[09-Sep-2011 13:06:56] <dhopp> you shouldn't login as admin :-P

[09-Sep-2011 13:07:05] <photon01> I guess so !

[09-Sep-2011 13:07:08] <dhopp> admin should be used to fix broken stuff

[09-Sep-2011 13:07:15] <photon01> <dhopp> : well, there's a way

[09-Sep-2011 13:07:38] <photon01> <dhopp> : move Data.fs.old to Data.fs

[09-Sep-2011 13:07:52] <photon01> <dhopp> : but I don't know how old is that .old thing

[09-Sep-2011 13:08:00] <dhopp> yeah you could probably do that and just lose any changes you have made

[09-Sep-2011 13:08:05] <dhopp> what's the timestamp on it?

[09-Sep-2011 13:08:06] <dhopp> ls -l

[09-Sep-2011 13:08:33] <photon01> let me get my VPN up

[09-Sep-2011 13:17:51] <jb>  any ent customers here?

[09-Sep-2011 13:18:04] <gwb235> a few

[09-Sep-2011 13:18:16] <jb> wondering if anyone got an email regarding 4.0.2 from rusty yesterday?

[09-Sep-2011 13:18:24] <jb> i think I did, but I accidentally deleted it.

[09-Sep-2011 13:18:42] <gwb235> was it the "we noticed you downloaded it" email?

[09-Sep-2011 13:33:28] <jmp242> Hey everyone - ZCA UPDATE!!!!!

[09-Sep-2011 13:33:31] <jmp242> community/zca/blog/2011/09/09/zca-update-for-september-9th-2011--zcaportalorg-launched

[09-Sep-2011 13:33:36] <jmp242> HOT off the presses

[09-Sep-2011 13:36:33] <SDuensin> jmp242, very hot!   "Apache 2 Test Page powered by CentOS"    :-)

[09-Sep-2011 13:36:58] <jmp242> interesting

[09-Sep-2011 13:37:02] <jmp242> that redirect was working

[09-Sep-2011 13:37:10] <jmp242> try zcaportal.org/wiki

[09-Sep-2011 13:38:33] <jmp242> let me update that lol

[09-Sep-2011 13:39:25] <jmp242> fixed now

[09-Sep-2011 13:39:30] <jmp242> in the post

[09-Sep-2011 13:42:08] <jmp242> sorry, we didn't put in the redirect from ZCAPortal.org to allow further updates for additional non-wiki stuff

[09-Sep-2011 13:44:39] <Hackman238> jmp242: Looks good. Thanks James!

[09-Sep-2011 13:44:59] <Hackman238> jmp242: I ended up subsidizing the SSL cert- couldn't get anyone to pick up the tab

[09-Sep-2011 13:45:36] <jmp242> Oh, we should have tried the free one then

[09-Sep-2011 13:45:40] <jmp242> but thanks

[09-Sep-2011 13:48:05] <illsci> where do you go to change the snmp community strings in 3.2.0

[09-Sep-2011 13:52:31] <jmp242> Configuration Properties

[09-Sep-2011 13:55:34] <stanchan> Infrastructure > Devices > Details > Configuration Properties if you can't find it

[09-Sep-2011 13:57:33] <jmp242> Oh, yea, if I say something incomprehensible, please let me know and I'll try and explain better

[09-Sep-2011 14:00:55] <Hackman238> jmp242: TY

[09-Sep-2011 14:01:24] <jmp242> Hackman238: no problem

[09-Sep-2011 16:04:20] <dpetzel> Ive got some servers that were added, deleted, and later re-added again (dont ask...). For some of them I see stuff like

[09-Sep-2011 16:04:22] <dpetzel> 2011-09-09 13:02:10,056 ERROR Zope.ZCatalog: uncatalogObject unsuccessfully attempted to uncatalog an object with a uid of /zport/dmd/Devices/Server/Windows/devices/deviceName/os/software/Windows Resource Kit Tools.

[09-Sep-2011 16:04:36] <dpetzel> any idea how I clean that up?

[09-Sep-2011 16:04:37] <JohnnyNOC> have you tried running zenfixit ?

[09-Sep-2011 16:05:02] <dpetzel> no, I wasnt entirly clear from the posts whhich scripts are 'safe' and which are not

[09-Sep-2011 16:05:08] <JohnnyNOC> i've had similar issues - someone pointed me to zenfixit.py

[09-Sep-2011 16:05:34] <rocket> dpetzel: its nothing that needs to be cleaned up

[09-Sep-2011 16:05:55] <rocket> it wont impact anything .. its just trying to remove something thats already been removed

[09-Sep-2011 16:05:58] <dpetzel> rocket: as in does not harm, or will eventually clear up? its accounts for a ton of log noise

[09-Sep-2011 16:06:17] <rocket> its noisy but harmless

[09-Sep-2011 16:06:27] <rocket> and a known issue

[09-Sep-2011 16:06:28] <dpetzel> ok cool, good to know

[09-Sep-2011 16:11:01] <nate1> good day all

[09-Sep-2011 16:11:26] <nate1> what would be the best way for me to setup monitoring of disk space on a specific server's logical drive?

[09-Sep-2011 16:11:47] <tvincent> is there way to hard code the speed of a interface within ZenOSS instead of SNMP?

[09-Sep-2011 16:12:14] <nate1> right now, there is a "Hard Disks" section which has 4 listed already, and you can see the graphs for each one. I just need to set up a thresh hold and notifications for them

[09-Sep-2011 16:12:19] <nate1> well, one in particular

[09-Sep-2011 16:20:35] <illsci> hey sorry... i stepped away if anyone answered me

[09-Sep-2011 16:20:37] <dpetzel> nate1: I think you can copy the template and set a threshold on the local copy

[09-Sep-2011 16:20:43] <illsci> where do you go to change the snmp community strings in 3.2.0

[09-Sep-2011 16:20:46] <dpetzel> IE navigate to components

[09-Sep-2011 16:20:53] <dpetzel> HardDisks, click the disk in question

[09-Sep-2011 16:21:01] <illsci> I tried to add my comm string in the wizard on install but I guess it didn't work out...

[09-Sep-2011 16:21:03] <dpetzel> from the Display Dropdown, hit template

[09-Sep-2011 16:21:11] <dpetzel> and there will be a Create Local Copy

[09-Sep-2011 16:21:28] <nate1> sorry just saw this

[09-Sep-2011 16:21:28] <dpetzel> on the local copy define a threshold

[09-Sep-2011 16:21:42] <nate1> dpetzel: how do I do that?

[09-Sep-2011 16:22:14] <dpetzel> nate1: which create the copy, or set the threshold?

[09-Sep-2011 16:22:40] <nate1> I just don't understand what you are saying I guess

[09-Sep-2011 16:22:52] <dpetzel> nate1: NP, goto the device in question

[09-Sep-2011 16:23:05] <nate1> it is a server under the section /server/windows

[09-Sep-2011 16:23:08] <dpetzel> goto components/harddisks as if you wanted to review the utilization of that hard disk

[09-Sep-2011 16:23:17] <nate1> there

[09-Sep-2011 16:23:30] <dpetzel> see the display drop down?

[09-Sep-2011 16:23:36] <nate1> yup

[09-Sep-2011 16:23:43] <nate1> ah

[09-Sep-2011 16:23:44] <nate1> template?

[09-Sep-2011 16:23:59] <dpetzel> yeah

[09-Sep-2011 16:24:01] <dpetzel> create local copy

[09-Sep-2011 16:24:09] <nate1> on the drive in question?

[09-Sep-2011 16:24:13] <dpetzel> correct

[09-Sep-2011 16:24:29] <nate1> created

[09-Sep-2011 16:24:31] <dpetzel> once the local copy is created

[09-Sep-2011 16:24:36] <dpetzel> click the link "FileSystem"

[09-Sep-2011 16:25:02] <nate1> where is that located?

[09-Sep-2011 16:25:06] <dpetzel> You should be brough to the page where you can edit the template that is no unique to that file system

[09-Sep-2011 16:25:16] <dpetzel> the link? Its the wordfilesystem in the first column

[09-Sep-2011 16:25:29] <dpetzel> sorry harddisk

[09-Sep-2011 16:25:33] <jmp242> random, now Opera likes the SSL cert on the site

[09-Sep-2011 16:25:40] <nate1> ok

[09-Sep-2011 16:25:43] <nate1> clicked on harddisk

[09-Sep-2011 16:25:44] <jmp242> I wonder if the registrar had to finish registering it

[09-Sep-2011 16:25:50] <nate1> brings up the monitoring templates page

[09-Sep-2011 16:25:51] <dpetzel> should bring you tht e monitoring templates page

[09-Sep-2011 16:25:58] <nate1> yup

[09-Sep-2011 16:26:01] <dpetzel> and under the harddisk template

[09-Sep-2011 16:26:03] <nate1> I am in device class view

[09-Sep-2011 16:26:06] <dpetzel> you have an entry for that hard disk right?

[09-Sep-2011 16:26:10] <nate1> it didnt bring up any device

[09-Sep-2011 16:26:13] <nate1> just the menu on the left

[09-Sep-2011 16:26:21] <dpetzel> k from the menu find harddisk

[09-Sep-2011 16:26:31] <nate1> if I expand "server/Windows"

[09-Sep-2011 16:26:33] <nate1> I see harddisk

[09-Sep-2011 16:26:42] <nate1> locally defined

[09-Sep-2011 16:26:44] <dpetzel> there should be entry under harddisk

[09-Sep-2011 16:26:55] <nate1> what do you mean under?

[09-Sep-2011 16:27:05] <dpetzel> can you expand harddisk?

[09-Sep-2011 16:27:08] <nate1> nope

[09-Sep-2011 16:27:33] <dpetzel> 1 sec, getting a phone call

[09-Sep-2011 16:27:37] <nate1> ok

[09-Sep-2011 16:28:11] <jmp242> All right everyone, have a good weekend. See you all monday

[09-Sep-2011 16:28:46] <nate1> see ya

[09-Sep-2011 16:30:56] <dpetzel> nate1: sorry about that, so back to it your on the monitoring templates page right?

[09-Sep-2011 16:31:09] <nate1> yres

[09-Sep-2011 16:31:22] <dpetzel> and on the left menu you find hardisk right?

[09-Sep-2011 16:31:33] <nate1> yes

[09-Sep-2011 16:31:38] <nate1> it does not have any sub sections

[09-Sep-2011 16:31:55] <dpetzel> hmmm

[09-Sep-2011 16:32:24] <dpetzel> if yo back to where you made the copy, does the button still say copy, or does it now say delete??

[09-Sep-2011 16:32:33] <nate1> when I look at the drive in the device, the button says remove local copy

[09-Sep-2011 16:32:36] <nate1> like it should exist

[09-Sep-2011 16:33:11] <dpetzel> when you click on the hardisk graphs, does it have any graphs?

[09-Sep-2011 16:33:18] <nate1> yes

[09-Sep-2011 16:33:23] <nate1> IO Bytes and HDD Space

[09-Sep-2011 16:33:28] <nate1> for every drive

[09-Sep-2011 16:33:38] <nate1> I mean, each drive I click on ahs the data

[09-Sep-2011 16:33:46] <dpetzel> I really dont understand how you can have a hard disk template under monitoring templates.....

[09-Sep-2011 16:33:49] <dpetzel> cant*

[09-Sep-2011 16:33:53] <dpetzel> im a bit stumped

[09-Sep-2011 16:33:54] <nate1> well

[09-Sep-2011 16:33:58] <nate1> I have a hard disk template

[09-Sep-2011 16:34:09] <nate1> called HardDisk (Locally Defined)

[09-Sep-2011 16:34:13] <nate1> it is under server/windows

[09-Sep-2011 16:34:22] <nate1> but under HardDisk, there is nothing

[09-Sep-2011 16:34:28] <dpetzel> ohh. are you sorted by Device class? button on the bottom?

[09-Sep-2011 16:34:31] <nate1> no other "harddisk" entries etc

[09-Sep-2011 16:34:36] <nate1> yes, device class

[09-Sep-2011 16:34:43] <dpetzel> ahh, im looking at it from "Template"

[09-Sep-2011 16:34:51] <nate1> ok

[09-Sep-2011 16:34:53] <nate1> now I am also

[09-Sep-2011 16:34:57] <nate1> I see two things under it

[09-Sep-2011 16:35:00] <nate1> /server/windows

[09-Sep-2011 16:35:07] <nate1> and /server/windows/devices/...

[09-Sep-2011 16:35:08] <nate1> etc.

[09-Sep-2011 16:35:14] <nate1> which I am guessing is the copy?

[09-Sep-2011 16:35:23] <dpetzel> the second entry is the full path thtehardisk in question right?

[09-Sep-2011 16:35:28] <dpetzel> to the hard disk*

[09-Sep-2011 16:35:53] <dpetzel> if, so yeah thats your copy

[09-Sep-2011 16:36:05] <dpetzel> you should be able to click it and edit/add thrshold information

[09-Sep-2011 16:36:55] <nate1> yes

[09-Sep-2011 16:36:58] <nate1> it appears so

[09-Sep-2011 16:37:00] <nate1> ok

[09-Sep-2011 16:37:02] <nate1> I will try that

[09-Sep-2011 16:37:03] <nate1> thanks!

[09-Sep-2011 16:37:06] <dpetzel> np

[09-Sep-2011 16:37:13] <nate1> it already has one up there

[09-Sep-2011 16:37:32] <nate1> I just have to add the threshold

[09-Sep-2011 16:37:41] <dpetzel> or edit the existing one

[09-Sep-2011 16:39:01] <nate1> now

[09-Sep-2011 16:39:04] <nate1> that I have this

[09-Sep-2011 16:39:11] <nate1> what is the best way to set up an alert

[09-Sep-2011 16:39:14] <nate1> for an item like this?

[09-Sep-2011 16:39:20] <nate1> the severity is set for warning

[09-Sep-2011 16:40:11] <nate1> ok, somethig seems wrong

[09-Sep-2011 16:40:24] <nate1> it appears that the changes I made effected all of the drives EXCEPT for the one I wanted it to

[09-Sep-2011 16:40:46] <nate1> system wide

[09-Sep-2011 16:40:52] <nate1> oh

[09-Sep-2011 16:40:55] <nate1> because I am an idiot

[09-Sep-2011 16:40:56] <nate1> lol

[09-Sep-2011 16:41:00] <nate1> I made changes to the wrong template

[09-Sep-2011 16:41:01] <dpetzel> sounds as though you edited the wrong template

[09-Sep-2011 16:41:22] <dpetzel> as for the alert, this threshold should create an event, so you just need to setup an alerting rule (see the admin guide for that)

[09-Sep-2011 16:41:57] <nate1> well I have setup rules for different thingsd

[09-Sep-2011 16:42:10] <nate1> let me see real quick

[09-Sep-2011 16:42:22] <nate1> I go to my user, and set up an alert

[09-Sep-2011 16:43:02] <nate1> and I set the action to email and enabled true and where to send it

[09-Sep-2011 16:43:24] <nate1> it's just I am not sure what filters to use to just grab that specific event

[09-Sep-2011 16:43:43] <dpetzel> yeah, there should be nothing special about the alert at this point, especially if you've done others before

[09-Sep-2011 16:45:26] <nate1> eh

[09-Sep-2011 16:45:29] <nate1> well

[09-Sep-2011 16:45:36] <nate1> I don't know what where criteria to use really:-p

[09-Sep-2011 16:45:45] <nate1> I can do device = <device name>

[09-Sep-2011 16:46:05] <nate1> rather

[09-Sep-2011 16:46:08] <nate1> Device is <name>

[09-Sep-2011 16:46:18] <nate1> not sure what filter to use after that though

[09-Sep-2011 16:49:30] <dpetzel> I dod component contains whatever drive letter I think

[09-Sep-2011 16:49:45] <dpetzel> although I've never filtered at the component level myself, so take that with a grain of salt

[09-Sep-2011 16:54:39] <nate1> I'll say component contains D

[09-Sep-2011 16:54:40] <nate1> lol

[09-Sep-2011 16:59:33] kocolosk is now known as kocolosk|away

[09-Sep-2011 17:39:23] <iruga> hi there

[09-Sep-2011 17:39:37] <iruga> somebody is monitoring extreme network devices?

[09-Sep-2011 19:45:42] kocolosk|away is now known as kocolosk

[09-Sep-2011 20:01:28] <photon01> Hi

[09-Sep-2011 20:01:31] <photon01> anyone here ?

[09-Sep-2011 20:35:12] kocolosk is now known as kocolosk|away

[09-Sep-2011 21:01:01] <willwh> photon01: yup

[09-Sep-2011 21:19:36] <photon01> <willwh> : resolved my problem, thanks

[09-Sep-2011 23:06:43] <willwh> photon01: np

[10-Sep-2011 00:00:01] [disconnected at Sat Sep 10 00:00:01 2011]

[10-Sep-2011 00:00:01] [connected at Sat Sep 10 00:00:01 2011]

[10-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[10-Sep-2011 12:00:01] [disconnected at Sat Sep 10 12:00:01 2011]

[10-Sep-2011 12:00:02] [connected at Sat Sep 10 12:00:02 2011]

[10-Sep-2011 12:00:15] [zenoss-logger (logger bot) has joined #zenoss]

[10-Sep-2011 15:52:33] bzed_ is now known as bzed

[11-Sep-2011 00:00:01] [disconnected at Sun Sep 11 00:00:01 2011]

[11-Sep-2011 00:00:01] [connected at Sun Sep 11 00:00:01 2011]

[11-Sep-2011 00:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[11-Sep-2011 11:37:58] <DW-10297> Does Zenoss work with twilio?

[11-Sep-2011 12:00:01] [disconnected at Sun Sep 11 12:00:01 2011]

[11-Sep-2011 12:00:02] [connected at Sun Sep 11 12:00:02 2011]

[11-Sep-2011 12:00:14] [zenoss-logger (logger bot) has joined #zenoss]

[11-Sep-2011 14:29:16] <DW-10297> Anyone know if there is a new zenpack for servertech/sentry CDUs that works with v3?

[11-Sep-2011 15:38:43] <DW-10297> Also does anyone know how to get Zenoss to stop freaking out when the number of httpd processes grows and shrinks dynamically?

[11-Sep-2011 15:42:31] <DW-10297> nevermind for some crazy reason it was watching two httpd processes one was http and one was /usr/sbin/httpd

[11-Sep-2011 17:07:37] kocolosk|away is now known as kocolosk

[11-Sep-2011 20:11:21] kocolosk is now known as kocolosk|away

[11-Sep-2011 20:12:47] <qkit> morning everyone

[11-Sep-2011 20:14:01] <qkit> guys, possible for me to set the /perf/systemfile disc space to anohter number? mean only trigger to another point that i specify

[11-Sep-2011 21:44:53] <Tabmow> How do people go about monitoring postgres for performance etc.? What packs are they using?

[11-Sep-2011 23:06:18] kocolosk|away is now known as kocolosk

[12-Sep-2011 00:00:01] [disconnected at Mon Sep 12 00:00:01 2011]

[12-Sep-2011 00:00:02] [connected at Mon Sep 12 00:00:02 2011]

[12-Sep-2011 00:00:14] [zenoss-logger (logger bot) has joined #zenoss]

[12-Sep-2011 03:07:01] <konvert> morning

[12-Sep-2011 05:35:41] <tsener> hello

[12-Sep-2011 07:57:54] <illsci> morning

[12-Sep-2011 08:03:33] klone is now known as psyklone

[12-Sep-2011 08:32:19] <konvert> morning

[12-Sep-2011 08:34:30] <psyklone> howdy

[12-Sep-2011 08:34:37] <illsci> I changed the zSnmpCommunities to my community string and I removed public and private and everytime I do a snmpwalk through zenoss it still uses public

[12-Sep-2011 08:34:47] <illsci> how do I get it to use my community string and remodel the devices

[12-Sep-2011 08:34:59] <illsci> do I have to restart something after you make changes to the zproperties or something

[12-Sep-2011 08:35:24] <konvert> normally you dont have to restart something

[12-Sep-2011 08:37:12] <psyklone> illsci, go look in the zProperty that's the next one down. zSnmpCommunity. that's the one you want to change

[12-Sep-2011 08:43:03] <Hackman238> psyklone: whats zSnmpCommunities used for?

[12-Sep-2011 08:43:12] <Hackman238> Morning all!

[12-Sep-2011 08:43:34] <psyklone> Hackman238, iterated through during discovery

[12-Sep-2011 08:44:03] <Hackman238> psyklone: ah

[12-Sep-2011 08:44:40] <konvert> Hackman238 hi

[12-Sep-2011 08:47:45] <jmp242> morning

[12-Sep-2011 08:54:09] <Sam-I-Am> moo.

[12-Sep-2011 08:59:29] <Hackman238> How's everyone?

[12-Sep-2011 08:59:46] <Sam-I-Am> trying to wake up

[12-Sep-2011 08:59:55] <Hackman238> I'm freaking beat from a long hike...thnik my legs are going to fall off

[12-Sep-2011 09:00:21] <Sam-I-Am> not sure why i bother sleeping... feel tired when i wake up anyway

[12-Sep-2011 09:00:26] <Sam-I-Am> the solution is caffeine

[12-Sep-2011 09:02:17] <jmp242> caffeine is your friend

[12-Sep-2011 09:02:18] <JohnnyNOC> ^--- agreed

[12-Sep-2011 09:02:39] <JohnnyNOC> though i've been thinking i should be looking for nuclear strength coffee beans

[12-Sep-2011 09:07:52] <Hackman238> JohnnyNOC: Plutonium beans

[12-Sep-2011 09:08:20] <Hackman238> JohnnyNOC: Don't have too many....have to prevent going critical mass

[12-Sep-2011 09:09:05] <illsci> How can I remodel devices I've already added or force zenoss to do another check of the hosts

[12-Sep-2011 09:09:15] <illsci> I just enabled icmp requests so now they should show up...

[12-Sep-2011 09:10:28] <Hackman238> Off topic, but we've got two Level 2 linux administrastor and two network security positions open. Anyone interested, let me know.

[12-Sep-2011 09:11:07] <Hackman238> illsci: Remodeling won't adavance the devices position in the zenping queue

[12-Sep-2011 09:11:38] <illsci> So you just have to wait?

[12-Sep-2011 09:11:50] <Hackman238> illsci: the devices aren't being pinged?

[12-Sep-2011 09:12:03] <illsci> well until a few seconds ago it wasn't working

[12-Sep-2011 09:12:15] <illsci> I didn't know how often it happens

[12-Sep-2011 09:13:01] <Hackman238> illsci: Zenoss should ping all the devices every 1 minute by default

[12-Sep-2011 09:13:09] <Hackman238> Zenping has a 1 minute cycle time.

[12-Sep-2011 09:13:18] <Hackman238> illsci: We can test to see if zenping is working

[12-Sep-2011 09:13:29] <illsci> I think it's working...

[12-Sep-2011 09:13:34] <illsci> I just saw some icmp packets come across

[12-Sep-2011 09:13:39] <Hackman238> illsci: Oh ok

[12-Sep-2011 09:14:07] <Hackman238> illsci: When a device is initially added a number of factors can contribute the the slow addition of the device to the polling.

[12-Sep-2011 09:14:41] <Hackman238> illsci: All in all, pinging should start within 1 to 3 zenping cycles, (1-3min) and the same often goes for the other daemons (5-15 minutes)

[12-Sep-2011 09:27:56] <konvert> is it possible to put a portlet on the dashboard with graphs of the filesystem from the devices with the filesystem higher than 90%

[12-Sep-2011 09:28:10] <konvert> i allready installed the graph portlet zenpack

[12-Sep-2011 09:29:55] <Sam-I-Am> no

[12-Sep-2011 09:30:06] <Sam-I-Am> all the graph portlet does is stuff graphs on the dashboard

[12-Sep-2011 09:30:19] <konvert> ok but can i make a graph report like that?

[12-Sep-2011 09:30:47] <Sam-I-Am> possibly, with some coding

[12-Sep-2011 09:31:00] <Sam-I-Am> why not just have an event pop up when a FS goes above 90%?

[12-Sep-2011 09:31:41] <konvert> i allready have this, but tought it would be good to have all filesystems above 90% on my dashboard

[12-Sep-2011 09:32:17] <Sam-I-Am> might be able to make a graph with a manual y axis that only shows stuff above 90%

[12-Sep-2011 09:35:04] <Hackman238> konvert: Sure

[12-Sep-2011 09:35:19] <Hackman238> konvert: BRB

[12-Sep-2011 09:39:12] <konvert> ok hackman can you try to talk me through?

[12-Sep-2011 09:40:54] <DW-10297> is there a way to make zenoss accept SNMP traps?

[12-Sep-2011 09:42:02] <konvert> i think zenoss automatically accepts snmp traps

[12-Sep-2011 09:42:25] <psyklone> DW-10297, the 'zentrap' daemon provides this functionality

[12-Sep-2011 09:42:54] <Hackman238> konvert: You'd need to create an event view that has the conditions for the 90% events

[12-Sep-2011 09:43:16] <Hackman238> konvert: Then using the event portlet, display that event view

[12-Sep-2011 09:43:52] <Hackman238> konvert: I can't recall if that pack work in v3. You'll have to try it in a vm or test area first less its confirmed working

[12-Sep-2011 09:47:11] <dpetzel> Hackman238: do I recall correctly that you had some sample reports where you could pass a time range, and report the max withing that time range?

[12-Sep-2011 09:47:30] <dpetzel> * from an RRD Datasource

[12-Sep-2011 09:47:59] <Hackman238> dpetzel: Fort he CLI? Yes

[12-Sep-2011 09:48:28] <dpetzel> Hackman238: are they somewhere I could look at them? I have a need to put some together and looking for a jump start

[12-Sep-2011 09:49:10] <Hackman238> dpetzel: message/60977#60977

[12-Sep-2011 09:49:32] <dpetzel> Hackman238: perfect thanks

[12-Sep-2011 09:50:06] <Hackman238> dpetzel: NP

[12-Sep-2011 09:54:42] <DW-10297> psyklone: is that something you setup in the Gui?

[12-Sep-2011 09:59:29] <DW-10297> ah okay I see it's running in advanced -> daemons

[12-Sep-2011 10:01:30] <psyklone> DW-10297, yeah, you've got it. it should be running by default so there's no need to do anything in the UI (typically) ... but if it's not running for one reason or another you can either start it from the GUI, or via 'zentrap start' at the CLI

[12-Sep-2011 10:08:10] <DW-10297> psyklone: so it will automatically accept traps for anything in 'infrastructure' or is there something you have to bind to the device?

[12-Sep-2011 10:10:07] <psyklone> DW-10297, it will accept traps for anything that you point to it. zenoss will create an event for a device that it receives a trap for, whether or not it's in your 'infrastructure'

[12-Sep-2011 10:13:06] <DW-10297> psyklone: hopefully final question, does zentrap need the community to be set to something specific in order for it to accept the trap?

[12-Sep-2011 10:14:47] <konvert> i'm still having trouble monitoring my mssql machine with the zenpack

[12-Sep-2011 10:14:54] <jmp242> Anyone use the netapp community zenpacks that could comment on this thread?

[12-Sep-2011 10:14:55] <jmp242> message/61303#61303

[12-Sep-2011 10:17:14] <Hackman238> jmp242: Sure.

[12-Sep-2011 10:23:51] <jmp242> Thanks Hackman238

[12-Sep-2011 10:24:17] <Hackman238> jmp242: No, TY for keeping an eye on the forums.

[12-Sep-2011 10:24:25] <jmp242> Hey, I do it anyway

[12-Sep-2011 10:24:55] <jmp242> We've had some registrations on the Wiki too, so good on that

[12-Sep-2011 10:25:08] <Hackman238> jmp242: Nice

[12-Sep-2011 10:25:46] <jmp242> The search is just waiting on a perl module that isn't in a repo

[12-Sep-2011 10:25:53] <jmp242> and I'll enable that once that's worked out this week

[12-Sep-2011 10:27:37] <Hackman238> jmp242: Sounds good. Thanks

[12-Sep-2011 10:28:52] <jmp242> I'm just glad to contribute

[12-Sep-2011 10:28:57] <DW-10297> Is it difficult to make Zenoss understand 'PowerMIB' traps?

[12-Sep-2011 10:29:26] <jmp242> DW-10297: it can be difficult generally for Zenoss to understand Traps... but I don't know that those are any more problematic than any other

[12-Sep-2011 10:29:27] <Hackman238> DW-10297: Its hard to make Zenoss understand any mibs

[12-Sep-2011 10:29:34] <konvert> someone some free time to help me trough this docs/DOC-3391

[12-Sep-2011 10:30:23] <jmp242> konvert: can you identify what in the docs is lacking so we can improve it?

[12-Sep-2011 10:30:28] <jmp242> i.e. where are you getting stuck

[12-Sep-2011 10:31:20] <konvert> in making the odbc driver

[12-Sep-2011 10:31:37] <konvert> i've installed all zenpacks, no trouble with that

[12-Sep-2011 10:32:27] <jmp242> Hmm, which driver did you decide to use?

[12-Sep-2011 10:32:31] <jmp242> the FreeDTS or the pyodbc?

[12-Sep-2011 10:32:33] <konvert> i think i succesfully installed freeTDS, but then i'm stuck

[12-Sep-2011 10:32:41] <konvert> for me either free one is good

[12-Sep-2011 10:33:04] <konvert> when i tried to install pyodbc it gives me an error

[12-Sep-2011 10:33:26] <jmp242> There isn't a permalink on the comments

[12-Sep-2011 10:33:33] <jmp242> but did you do what this comment says

[12-Sep-2011 10:33:33] <jmp242> Snake-BIS Sep 30, 2010 6:13 AM

[12-Sep-2011 10:33:37] <jmp242> ?

[12-Sep-2011 10:33:41] <jmp242> for FreeDTS

[12-Sep-2011 10:35:35] <konvert> jmp242 i'm using suse-linux so apt-get doesn't work

[12-Sep-2011 10:48:04] <konvert> and i dont find the conf file : odbcinst.ini

[12-Sep-2011 11:00:26] <Jane_Curry> Hiya

[12-Sep-2011 11:00:54] <Jane_Curry> Could anyone do a quick confirm / deny for me on complex community names for cisco kit

[12-Sep-2011 11:01:15] <Jane_Curry> I have a box with a community name of fred@jane

[12-Sep-2011 11:01:34] <Jane_Curry> Do I just put fred@jane in the zSnmpCommunity property?

[12-Sep-2011 11:01:43] <Jane_Curry> Does it need quoting or escaping

[12-Sep-2011 11:02:03] <DW-10297> after you import mibs do you need to restart?

[12-Sep-2011 11:02:23] <Jane_Curry> As you have no doubt guessed, zenmodeler isn't getting any response but an snmpwalk succeeds from a command line

[12-Sep-2011 11:02:38] <psyklone> DW-10297, restart zentrap

[12-Sep-2011 11:03:09] <psyklone> Jane_Curry, hmmm ... that's a good question. i'll have a look after i refresh my coffee

[12-Sep-2011 11:06:39] <Jane_Curry> Thanks psyklone - I need to pop out for an hour but any input would be much appreciated

[12-Sep-2011 11:07:34] kocolosk is now known as kocolosk|away

[12-Sep-2011 11:07:59] <Jane_Curry> An snmpwaqlk from the Commands popup DOES work and you can see the fred@jane community - it's just the modelers that hang

[12-Sep-2011 11:13:03] <bigegor> konvert: as zenoss user run: easy_install-2.6 pyodbc==2.1.8

[12-Sep-2011 11:19:33] <konvert> bigegor i get error, gcc: error trying to exec 'cclplus':execvp: Permission denied

[12-Sep-2011 11:21:08] <DW-10297> if zenmib finds 1 MIBs to import but then says Loaded 0 MIB files anyone know what the deal is on that?

[12-Sep-2011 11:26:54] <konvert> bigegor ?

[12-Sep-2011 11:30:29] <bigegor> konvert: your error is destributions specific. Ask google. Note, that you must also install some dependencies (odbc-dev, g++)

[12-Sep-2011 11:38:45] <psyklone> Jane_Curry, i just took a look at 2.5.2, and it passes the comm string into the modeler without needing to escape '@'. when i run zenmodeler against my linux server that i've set the string to 'test@public' on, i can see that in the debug output it's sending the string intact and it's collecting the data as expected

[12-Sep-2011 11:40:41] <Bart> I'm using @s in community strings to monitor juniper devices, works fine without escaping. Also on 2.5.2, i hope its not broken in v3, all my routers need this

[12-Sep-2011 11:43:40] <psyklone> Bart - just checked on 3 to be sure, works fine

[12-Sep-2011 11:47:40] <dhopp> hrm…if I'm modeling a windows machine and I want it to ignore all filesystems other than "c:" what do I set zFileSystemMapIgnoreNames to? I can't seem to get it to work

[12-Sep-2011 11:49:04] <JohnnyNOC> D|E|F|G|H

[12-Sep-2011 11:49:10] <JohnnyNOC> and so on (seems to work for me)

[12-Sep-2011 11:50:14] <psyklone> i'd imagine you could just do ^[^c]

[12-Sep-2011 11:50:33] <dhopp> psyklone:  I literally just tried that :-)

[12-Sep-2011 11:50:35] <dhopp> and it worked

[12-Sep-2011 11:50:47] <psyklone> woot

[12-Sep-2011 11:50:59] <dhopp> JohnnyNOC: yeah I knew that could work..but regex should be cooler than that :-P

[12-Sep-2011 11:51:28] <JohnnyNOC> gotcha

[12-Sep-2011 11:51:31] <dhopp> psyklone:  I was just getting ready to type "nevermind I got it with ^[^c]..hehe

[12-Sep-2011 12:00:01] [disconnected at Mon Sep 12 12:00:01 2011]

[12-Sep-2011 12:00:01] [connected at Mon Sep 12 12:00:01 2011]

[12-Sep-2011 12:00:15] [zenoss-logger (logger bot) has joined #zenoss]

[12-Sep-2011 12:25:33] <nate1> Hello all

[12-Sep-2011 12:25:58] <nate1> I am getting some annoying notifications from Zenoss

[12-Sep-2011 12:26:17] <nate1> long astory short... I set up a threshold on a D Hard DIsk on a server

[12-Sep-2011 12:26:41] <nate1> and then set up an alerting rule to send a text whenever the threshold was exceeded

[12-Sep-2011 12:27:21] <nate1> on the rule, I have it set to where Device is "server name"

[12-Sep-2011 12:27:26] <nate1> and component is "D"

[12-Sep-2011 12:27:29] <nate1> which is the drive

[12-Sep-2011 12:28:15] <nate1> but, every few hours I have been getting text notifications with the custom title that I set for this alert

[12-Sep-2011 12:28:24] <nyeates> 3.2.1 maint release being prepped for a beta release to community

[12-Sep-2011 12:28:46] <nate1> and the message says "MSN: Error reading value from <name> on... etc

[12-Sep-2011 12:29:40] <nate1> there are 4 texts at the same time for "TotalbytesSent_LowWord", "TotalbytesSent_HighWord", "MeasuredBandwidth", and "IDiskInstance"

[12-Sep-2011 12:29:49] <nate1> I do not see any events for those..

[12-Sep-2011 12:29:57] <nate1> yet I am getting spam texted about them

[12-Sep-2011 12:29:58] <nate1> any ideas?

[12-Sep-2011 12:32:28] <dhopp> nyeates:  very cool

[12-Sep-2011 12:32:45] <dhopp> nyeates:  is it just those two bug fixes?

[12-Sep-2011 12:33:25] <jmp242> nate1: it may be that windows isn't returning that data for some reason

[12-Sep-2011 12:33:46] <nate1> which data?

[12-Sep-2011 12:33:50] <nate1> those 4 values?

[12-Sep-2011 12:34:12] <nate1> the thing is... those 4 values are in "Device" under monitoring templates

[12-Sep-2011 12:34:25] <nate1> not even related to the HardDisk monitoring template

[12-Sep-2011 12:34:43] <nate1> so I don't know why it would be catching those errors and sending them

[12-Sep-2011 12:35:16] <jmp242> Check your alerting rule definition then

[12-Sep-2011 12:35:26] <jmp242> you may have to filter more specifically

[12-Sep-2011 12:35:47] <Jane_Curry> Thanks psyklone - appreciate that

[12-Sep-2011 12:35:53] <psyklone> np

[12-Sep-2011 12:36:14] <nate1> hmm

[12-Sep-2011 12:36:26] <nate1> those messages don't even appear under events

[12-Sep-2011 12:36:32] <nate1> I will change the threshold to throw an info

[12-Sep-2011 12:36:38] <nate1> and add that as an alerting filter

[12-Sep-2011 12:36:41] <nate1> and see what that does

[12-Sep-2011 12:37:51] <dhopp> nate1: are you sure they aren't being moved to history right away?

[12-Sep-2011 12:37:57] <Jane_Curry> nyeates - do you have an ETA?? I was planning on doing a first install of 3.2 tomorrow...

[12-Sep-2011 12:38:15] <nate1> is there a way I can see history?

[12-Sep-2011 12:38:31] <dhopp> when you are looking at the event console there is a history button/link

[12-Sep-2011 12:40:09] <nate1> hmm

[12-Sep-2011 12:40:12] <nate1> don't see them in history

[12-Sep-2011 12:50:36] <jmp242> mmm

[12-Sep-2011 12:50:49] <jmp242> That's really odd for it to generate an alert without an event to see even in History

[12-Sep-2011 12:51:03] <jmp242> are you only getting downs, or also CLEAR alerts?

[12-Sep-2011 12:51:14] <nate1> just down

[12-Sep-2011 12:51:15] <nate1> s

[12-Sep-2011 12:51:20] <nate1> or rather the error msg

[12-Sep-2011 12:51:28] <nate1> I changed it though

[12-Sep-2011 12:51:34] <nate1> so hopefully I shouldn't receive them

[12-Sep-2011 12:51:38] <nate1> will see if I get anymore today

[12-Sep-2011 12:51:48] <nate1> and if I do, I will get back:)

[12-Sep-2011 13:05:03] kocolosk|away is now known as kocolosk

[12-Sep-2011 13:12:16] kocolosk is now known as kocolosk|away

[12-Sep-2011 13:12:40] kocolosk|away is now known as kocolosk

[12-Sep-2011 13:15:10] <jmp242> Anyone else get this was deleted on this thread?

[12-Sep-2011 13:15:11] <jmp242> message/61317#61317

[12-Sep-2011 13:15:24] <jmp242> I can't figure out why it would have been deleted really . . .

[12-Sep-2011 13:19:36] <dpetzel> jmp242: Confirmed I also get it may have been deleted message

[12-Sep-2011 13:22:13] <Hackman238> jmp242: Weird

[12-Sep-2011 13:22:28] <Hackman238> Didnt I look at that one earlier?

[12-Sep-2011 13:23:14] <Hackman238> Nope guess not. IDK, looks like its gone

[12-Sep-2011 13:26:48] <jmp242> Strange

[12-Sep-2011 13:26:49] <jmp242> I

[12-Sep-2011 13:26:57] <jmp242> I'd been helping him with running zenfixit

[12-Sep-2011 13:27:07] <jmp242> oh well

[12-Sep-2011 13:27:19] <jmp242> nyeates: it wasn't administratively deleted was it?

[12-Sep-2011 13:42:14] <nyeates> just seeing comments to me here....let me catch up in a sec

[12-Sep-2011 13:49:54] <illsci> hey what were those two zenpatches I need for 3.2.0?  the fix snmp modeling stuff

[12-Sep-2011 13:50:00] <illsci> how do you find the available zenpatches?

[12-Sep-2011 13:50:23] <nyeates> illsci: look in the subject

[12-Sep-2011 13:50:30] <illsci> got it

[12-Sep-2011 13:56:10] <DW-10297> is there a way to show all graphs from a device class on the dashboard?

[12-Sep-2011 13:59:15] <jmp242> DW-10297: not that I'm aware of, though maybe something like the site window portlet to a multigraph report?

[12-Sep-2011 14:00:03] <DW-10297> ah there is a graph portlet i guess if I can figure out how to install a new portlet =D

[12-Sep-2011 14:08:30] <DW-10297> If you create a multi-graph report and just put a whole device class into a collection is that all you need to do?

[12-Sep-2011 14:08:44] <DW-10297> I'm basically just trying to get one screen to show all of the graphs for everything in /power

[12-Sep-2011 14:09:45] <nyeates> dhopp: 3.2.1 will include the two big bugs that community found, in addition to 2 other possible ones, or anything else that comes up.

[12-Sep-2011 14:11:33] crazed_ is now known as crazed

[12-Sep-2011 14:11:43] <nyeates> Jan_Curry: there is not an ETA that I have heard yet. I do know that we want to do a 3.2.1 community beta - just to make sure - and then realse it. I would say 2 weeks out?

[12-Sep-2011 14:12:12] <nyeates> Jane_Curry, is what I had meant (in case the proper spelling of your name flags you)

[12-Sep-2011 14:15:41] <Jane_Curry> nyeates - OK - I'll go ahead and install base 3.2 then, plus patches.  Are there just 2 still?

[12-Sep-2011 14:16:26] <nyeates> Yes, just two that I am aware of ....other issues are being replicated and tested still

[12-Sep-2011 14:16:43] <nyeates> various catalogue problems that we havent been able to repeat in house

[12-Sep-2011 14:17:27] <davetoo> nick I didn't know you were canadian

[12-Sep-2011 14:17:35] <DW-10297> Is there an easy way to create a multi-graph report for everything under /power ?

[12-Sep-2011 14:17:36] <jb> afternoon.

[12-Sep-2011 14:17:37] <dhopp> Jane_Curry: you might to get zenpatch 26336…it doesn't fix anything, but fixes an annoying display when looking at modeler plugins

[12-Sep-2011 14:17:50] <dhopp> err..fix anything that is technically broken I mean

[12-Sep-2011 14:17:59] <nyeates> davetoo: is catalogue the canadian spelling or something?

[12-Sep-2011 14:18:13] <Hackman238> nyeates: UK

[12-Sep-2011 14:18:30] <davetoo> that

[12-Sep-2011 14:18:40] <nyeates> I guess I should tell you the colour of the programme either

[12-Sep-2011 14:18:44] <davetoo> the non-Amerikan spelling

[12-Sep-2011 14:19:48] <nyeates> dhopp: oh shnap. I didnt realize the modeler plugins had a zenpatch already....thx

[12-Sep-2011 14:20:21] <nyeates> yeah modeler plugins were reversed (left to right) and not labeled... that fix likely changes that!

[12-Sep-2011 14:21:11] <nyeates> That will be included in 3.2.1 also

[12-Sep-2011 14:21:25] <psyklone> nyeates, that fix addresses the column titles. the left/right switch was left in place to be uniform with the way that things are done elsewhere in the UI

[12-Sep-2011 14:25:12] fOrsberg is now known as forsberg

[12-Sep-2011 14:32:32] <DW-10297> When you create a multi-graph report how do you specify the name of the object you want graphed?

[12-Sep-2011 14:35:36] <DW-10297> nevermind you can add multiple graphs to a single graph report..

[12-Sep-2011 14:41:45] <DW-10297> Can zenoss send reports via email?

[12-Sep-2011 14:59:51] <Hackman238> DW-10297: Yes, reportmail

[12-Sep-2011 15:06:22] <bender183> ah the famous Jane_Curry, hello

[12-Sep-2011 15:11:42] <DW-10297> is there any way to formulate the URL when passing it to report mail so that it will send an email that makes sense?

[12-Sep-2011 15:11:55] <DW-10297> all it seems to send is the text from the report and not the images

[12-Sep-2011 15:15:50] <DW-10297> hmm yea reportmail seems pretty terrible

[12-Sep-2011 15:16:25] <jmp242> mmm

[12-Sep-2011 15:16:35] <jmp242> Sorry I don't use it

[12-Sep-2011 15:16:56] <DW-10297> wonder if I can script it with curl

[12-Sep-2011 15:16:59] <DW-10297> probably work better

[12-Sep-2011 15:20:14] <jmp242> could be

[12-Sep-2011 15:21:02] <DW-10297> it's not really that hard to grab the body of an HTML document and email it to someone =)

[12-Sep-2011 15:21:07] <DW-10297> not sure why reportmail can't do it.

[12-Sep-2011 15:27:51] <Jane_Curry> bender183 - maybe it's the INfamous Jane Curry

[12-Sep-2011 16:01:59] <uniplexed> hi there

[12-Sep-2011 16:02:11] <uniplexed> is there any way to remodel all devices in one step?

[12-Sep-2011 16:02:50] <jmp242> Zenmodeler will run every 12 hours

[12-Sep-2011 16:03:02] <jmp242> You could probably just run that youself

[12-Sep-2011 16:03:08] <jmp242> check the cli arguments

[12-Sep-2011 16:03:46] <psyklone> `zenmodeler run`will invoke a modeling job against everything

[12-Sep-2011 16:03:47] <jmp242> probably zenmoder run

[12-Sep-2011 16:03:51] <Adam_z> zenmodeler run does it

[12-Sep-2011 16:03:54] <psyklone> jinx!

[12-Sep-2011 16:04:12] <jmp242> I had to check the cli arguments

[12-Sep-2011 16:04:19] <jmp242> I never want to run a full zenmodel manually so . . .

[12-Sep-2011 16:05:36] <forsberg> is it possible to get a list of manageip on all devices in a deviceclass? zendmd maybe

[12-Sep-2011 16:06:51] <uniplexed> thanks

[12-Sep-2011 16:06:55] <uniplexed> i just update zenoss to 3.2

[12-Sep-2011 16:06:58] <uniplexed> and i need to remodel

[12-Sep-2011 16:07:00] <uniplexed> thanks all

[12-Sep-2011 16:07:02] <psyklone> forsberg, yep, zendmd would be the ticket

[12-Sep-2011 16:28:06] <uniplexed> has someone used fping from blake in 3.2?

[12-Sep-2011 16:29:03] <jmp242> uniplexed: you need to install the patches

[12-Sep-2011 16:29:16] <jmp242> or wait for 3.2.1 - there are command based datasources that don't work right in 3.2

[12-Sep-2011 16:32:30] <uniplexed> i see

[12-Sep-2011 16:32:33] <uniplexed> i reinstalled the zenpacks

[12-Sep-2011 16:32:37] <uniplexed> but they dont seem to work

[12-Sep-2011 16:32:42] <uniplexed> fping and showportlets

[12-Sep-2011 16:32:50] <uniplexed> i was using them on 3.1

[12-Sep-2011 16:32:54] <jmp242> Showportlets I don't know

[12-Sep-2011 16:33:03] <jmp242> but fping you probably need to do the patches - see the topic

[12-Sep-2011 16:39:04] <uniplexed> where can i get those patches??

[12-Sep-2011 16:40:20] <Hackman238> uniplexed: use zenpatch

[12-Sep-2011 16:55:51] <bender183> Jane_Curry, I had a feeling you would say something like that

[12-Sep-2011 17:01:22] <dhopp> hrm…I have a device that is in a device class that is outside of the 'normal' device classes.  I have modeler plugins 'zenoss.snmp.NewDeviceMap', 'zenoss.snmp.DeviceMap', 'zenoss.snmp.HRFileSystemMap' and it models the device and picks up the filesystems.  However, no templates are being added to the file systems that are found and hence I get no graphs

[12-Sep-2011 17:01:23] <dhopp> ideas?

[12-Sep-2011 17:14:32] <dhopp> nevermind

[12-Sep-2011 17:14:45] <dhopp> the component template wasn't bound at any level that would allow it to be inherited

[12-Sep-2011 17:17:21] <uniplexed> thanks Hackman238

[12-Sep-2011 17:20:27] <uniplexed> i updated it, but i keep on receiving error collecting ethernet port data via snmp

[12-Sep-2011 17:20:38] <uniplexed> i applied enpatch 26334 and 35

[12-Sep-2011 17:21:11] <uniplexed> at least fping started to work again

[12-Sep-2011 17:26:40] <uniplexed> weird stuff, but it seems to work

[12-Sep-2011 18:10:15] <fus10nx> How can I set the rate at which emails are sent out for something being down i.e. keep sending them every 'x' minutes until acknowledged

[12-Sep-2011 18:14:09] <dhopp> fus10nx:  go to your alerting rule

[12-Sep-2011 18:14:18] <dhopp> and there is a "Repeat time (secs)"

[12-Sep-2011 19:09:26] kocolosk is now known as kocolosk|away

[12-Sep-2011 19:23:02] forsberg is now known as fOrsberg

[13-Sep-2011 00:00:01] [disconnected at Tue Sep 13 00:00:01 2011]

[13-Sep-2011 00:00:02] [connected at Tue Sep 13 00:00:02 2011]

[13-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[13-Sep-2011 02:33:19] <konvert> morning

[13-Sep-2011 03:30:10] fOrsberg is now known as forsberg

[13-Sep-2011 04:37:27] <konvert> hello someone awake yet?

[13-Sep-2011 04:51:53] <forsberg> slightly

[13-Sep-2011 04:52:17] * Simon4 is wide awake and bushy tailed

[13-Sep-2011 04:52:23] <Simon4> coffee #2 is kicking in

[13-Sep-2011 05:08:07] <konvert> nice, someone uses mssql monitoring?

[13-Sep-2011 05:08:23] <Simon4> not me sorry, all unix-style here

[13-Sep-2011 05:08:56] <konvert> and do u know how to install pyodbc on a linux machine?

[13-Sep-2011 05:09:49] <Simon4> sorry, other than googling

[13-Sep-2011 05:10:18] <forsberg> i do konvert

[13-Sep-2011 05:10:25] <forsberg> oh sorry ny

[13-Sep-2011 05:10:26] <forsberg> no

[13-Sep-2011 05:10:28] <forsberg> not mssql lol

[13-Sep-2011 05:19:42] <forsberg> anyone awake with some zendmd skills?

[13-Sep-2011 05:22:12] <forsberg> found out how to make a list of devices

[13-Sep-2011 05:22:14] <forsberg> for dev in dmd.Devices.getSubDevices():

[13-Sep-2011 05:22:14] <forsberg> dev

[13-Sep-2011 05:22:43] <forsberg> i wonder where one would defince a deviceclass, to only list devices in that class

[13-Sep-2011 05:27:00] <forsberg>  for dev in dmd.Devices.Server.Linux.Hosting.getSubDevices():

[13-Sep-2011 05:27:02] <forsberg> ahh

[13-Sep-2011 05:29:56] <konvert> so forsberg you are monitoring mssql machines?

[13-Sep-2011 07:40:09] <konvert> hello someone awake yet?

[13-Sep-2011 07:51:54] <Jane_Curry> hi

[13-Sep-2011 07:53:02] <konvert> hi Jane_Curry

[13-Sep-2011 07:53:10] <konvert> some time to help me?

[13-Sep-2011 07:53:30] <konvert> i'm struggling almost a week to get my mssql machines monitored with odbc

[13-Sep-2011 07:55:47] <Jane_Curry> so what have you done so far?  and what version of Zenoss and ZenPacks?

[13-Sep-2011 08:01:14] <konvert> zenoss 3.2 with the 2 patches

[13-Sep-2011 08:02:12] <konvert> zenpack : community.MsSQLMon_ODBC 3.1 with all requisite zenpacks

[13-Sep-2011 08:02:41] <konvert> the installation of the zenpacks is not the problem i think.

[13-Sep-2011 08:02:55] <konvert> My problem is when it comes to make the odbc connection

[13-Sep-2011 08:03:28] <konvert> i've installed freetds on the zenoss server, and tried to install pyodbc (think it worked to because i get no errors)

[13-Sep-2011 08:03:41] <konvert> And here's the point i get stuck

[13-Sep-2011 08:08:13] <Jane_Curry> what version of MS SQL are you trying to monitor?

[13-Sep-2011 08:10:56] <konvert> 2008

[13-Sep-2011 08:11:35] <Jane_Curry> Have you setup zMsSqlConnectionString zWinUser and zWinPassword as documented in the ZenPack page on zenoss.org?

[13-Sep-2011 08:11:45] <konvert> yes

[13-Sep-2011 08:12:03] <konvert> but i think this isn't enough,

[13-Sep-2011 08:12:49] <Jane_Curry> Is the community.odbc.MsSqlDatabaseMap modler plugin active for these devices?

[13-Sep-2011 08:12:55] <konvert> DRIVER={FreeTDS};TDS_Version=7.2;PORT=1433;UID=*****;PWD=*****

[13-Sep-2011 08:13:04] <konvert> yes

[13-Sep-2011 08:13:40] <Jane_Curry> Have you completed the zMsSqlSrvInstances property?

[13-Sep-2011 08:14:25] <konvert> yes, i just have to put my instance name in there?

[13-Sep-2011 08:15:01] <Jane_Curry> Try running zenmodeler with verbose logging for the community.odbc.MsSqlDatabaseMap modeler against your device - eg.

[13-Sep-2011 08:15:11] <Jane_Curry> As the zenoss user..

[13-Sep-2011 08:16:02] <Jane_Curry> zenmodeler run -v 10 -d <device name known to Zenoss> --collect MsSqlDatabaseMap

[13-Sep-2011 08:21:46] <konvert> done that

[13-Sep-2011 08:21:56] <konvert> what do i have to see there

[13-Sep-2011 08:23:37] <konvert> i've also read i need to try to connect with tsql -> works

[13-Sep-2011 08:24:06] <konvert> and then try to connect with isql -> doesn't work, or my syntax isn't correct

[13-Sep-2011 08:25:19] <Jane_Curry> Do you see any errors in the ouput from zenmodeler?

[13-Sep-2011 08:25:23] <konvert> do i need to post the outcome ?

[13-Sep-2011 08:25:43] <Jane_Curry> Sounds like the issue is with actually connecting to the database?

[13-Sep-2011 08:27:06] <konvert> yes i get error

[13-Sep-2011 08:28:21] <konvert> ERROR zen.SQLCLIENT:Received error (('00000', '[ISQL]ERROR: Could not SQLConnect')) from query: sp_helpdb

[13-Sep-2011 08:32:40] <konvert> do i need to make a DSN on the mssql machine? or do i need to do it on my zenoss machine

[13-Sep-2011 08:34:52] <Jane_Curry> Soory - I am no database guru and it sounds like that is what you need!

[13-Sep-2011 08:35:19] <Jane_Curry> Have you ploughed through the comments on the ZenPack page?

[13-Sep-2011 08:35:51] <konvert> yes, but i think my problem is i don't have the odbc.ini and odbcinst.ini file

[13-Sep-2011 08:40:16] <konvert> no i find the files, they are in etc/unixodbc/ but they are empty

[13-Sep-2011 08:41:39] <konvert> and the folder ODBCDataSources in etc/unixodbc is empty

[13-Sep-2011 08:45:09] <Jane_Curry> I'm sorry - you are way beyond my database knowledge here!

[13-Sep-2011 08:46:00] <konvert> thx anyway

[13-Sep-2011 08:46:26] <konvert> hopefully someone else can help me because this is taking way to long, and pisses me off

[13-Sep-2011 08:48:08] <Hackman238> konvert: whats up?

[13-Sep-2011 08:48:18] <konvert> hi Hackman238

[13-Sep-2011 08:48:38] <konvert> still no breaking through in my mssql monitoring setup

[13-Sep-2011 08:49:02] <konvert> have you read the converstion with Jane_Curry

[13-Sep-2011 08:49:16] <Hackman238> konvert: No, give me 5 to read it

[13-Sep-2011 08:49:34] <konvert> u can have 10

[13-Sep-2011 08:59:02] <Hackman238> konvert: is the DSN defined in /home/zenoss/odbc.ini?

[13-Sep-2011 08:59:45] <konvert> don't have that file there

[13-Sep-2011 09:00:07] <konvert> only have odbc.ini in unixodbc in /etc but that file is empty

[13-Sep-2011 09:07:45] <Hackman238> konvert: Okay you'll need to define the DNS in there *I think*

[13-Sep-2011 09:08:02] <Hackman238> konvert: Unfortunately bigegor is the master of the DB stuff

[13-Sep-2011 09:08:51] <konvert> and do u know what i have to put in there ?

[13-Sep-2011 09:09:03] <Hackman238> konvert: Theres a pretty extensive thread here about this problem: docs/DOC-3391#comment-4448

[13-Sep-2011 09:11:05] <Hackman238> konvert: Should be something like:

[13-Sep-2011 09:11:08] <Hackman238> [ms-sql]

[13-Sep-2011 09:11:08] <Hackman238> Description = TDS connection

[13-Sep-2011 09:11:08] <Hackman238> Driver = /usr/lib/libtdsodbc.so

[13-Sep-2011 09:11:08] <Hackman238> Setup = /usr/lib/libtdsS.so

[13-Sep-2011 09:11:08] <Hackman238> UsageCount = 1

[13-Sep-2011 09:11:12] <Hackman238> FileUsage = 1

[13-Sep-2011 09:11:43] <Hackman238> konvert: You'll also need to setup /etc/freetds.ini and /etc/odbcinst.ini

[13-Sep-2011 09:12:03] <Hackman238> konvert: I think all these areas are covered on that thread

[13-Sep-2011 09:12:57] <konvert> ok will try to look in that

[13-Sep-2011 09:31:19] <Sam-I-Am> sup guys

[13-Sep-2011 09:32:32] <Hackman238> Sam-I-Am: Hey

[13-Sep-2011 09:34:12] kocolosk|away is now known as kocolosk

[13-Sep-2011 09:45:34] <Hackman238> konvert: Any luck?

[13-Sep-2011 09:46:09] <konvert> hackman i've just seen some graphs coming up so YES

[13-Sep-2011 09:46:41] <konvert> thx for all the help

[13-Sep-2011 09:46:50] <konvert> you two Jane_Curry

[13-Sep-2011 09:50:01] forsberg is now known as fOrsberg

[13-Sep-2011 09:50:12] <konvert> another question, is it already possible to monitor exchange 2010 ?

[13-Sep-2011 09:54:08] <SDuensin> Morning.

[13-Sep-2011 10:00:23] <bender183> morning

[13-Sep-2011 10:01:04] <bender183> konvert, as long as you can enumerate snmp information from exchange 2010, you can monitor it

[13-Sep-2011 10:09:39] <konvert> but i mean if there is a zenpack for it

[13-Sep-2011 10:12:21] <bender183> community/zenpacks

[13-Sep-2011 10:20:11] <SEJeff_work> For existing ZenOSS Enterprise 3.x customers, am I reading right that we should wait for 4.2 to upgrade to ZSD?

[13-Sep-2011 10:20:21] <SEJeff_work> I'm updating from 3.1.0 ---> 3.2.0 now

[13-Sep-2011 10:23:39] <bender183> where are you reading that

[13-Sep-2011 10:25:02] <konvert> can someone tell me why i have a component databases on one device and on the other device i don't have it, 2 sql machines both same template and same modelers

[13-Sep-2011 10:25:36] <rhoulihan> you can upgrade to 3.2 now, ZSD does not come to core until 4.2

[13-Sep-2011 10:25:54] <rhoulihan> there will be an upgrade path from 3.2 to 4.2

[13-Sep-2011 10:26:21] <SEJeff_work> bender183, In the enterprise support portal

[13-Sep-2011 10:26:38] <SEJeff_work> When clicking "Zenoss Service Dynamics" on the left hand download bar there is only a README that says that

[13-Sep-2011 10:26:58] <SEJeff_work> rhoulihan, I'm an enterprise customer trying to get a copy to play with in our lab

[13-Sep-2011 10:27:09] <SEJeff_work> But if it isn't really GA yet then I'll just wait

[13-Sep-2011 10:27:20] <rhoulihan> ok, not a problem...those artifacts should be available for you

[13-Sep-2011 10:27:50] <rhoulihan> it is GA

[13-Sep-2011 10:27:58] <rhoulihan> as of 8/15

[13-Sep-2011 10:28:02] <gwb235> SEJeff_work: I downloaded the vmware.iso for 4.0.2 and attached it; the install went fairly smoothly but haven't played with it much, yet

[13-Sep-2011 10:28:22] <rhoulihan> 4.0.2 is the current release

[13-Sep-2011 10:28:50] <gwb235> for testing new releases I really appreciate the vmware packaging

[13-Sep-2011 10:28:54] <rhoulihan> actually GA was 7/15

[13-Sep-2011 10:29:14] <SEJeff_work> rhoulihan, Ah, well why doesn't someone update the portal with rpms? I'm installing this on a centos server and don't much care for a vmware iso

[13-Sep-2011 10:29:46] <rhoulihan> good question, let me find out for you.  one sec

[13-Sep-2011 10:29:55] <SEJeff_work> rhoulihan, thanks!

[13-Sep-2011 10:30:15] <SEJeff_work> Now if zenoss was provided as a qcow2 (kvm) image, that would be cool, but we use kvm and not vmware for virtualization

[13-Sep-2011 10:31:29] <gwb235> i'm with SEJeff_work; i'm looking at service dynamics as a green field implementation from our aging 2.5.2 infrastructure so downloaded the "latest" available (4.0.2).  If 4.2 was out I'd start with evaluating that.

[13-Sep-2011 10:33:06] <rhoulihan> got it...sounds like SEJeff is looking for an eval

[13-Sep-2011 10:34:13] <rocket> SEJeff_work: open a ticket in the portal requesting an eval of it, we have a few extra hoops with oracle licensing in 4.0 at the moment

[13-Sep-2011 10:34:24] <Hackman238> konvert: Glad you're up and running

[13-Sep-2011 10:34:27] <rocket> SEJeff_work: we are working to remove that particular roadblock

[13-Sep-2011 10:34:38] <SEJeff_work> rocket, Oh for the proprietary mysql

[13-Sep-2011 10:34:41] <SEJeff_work> Don't sweat it

[13-Sep-2011 10:34:49] <SEJeff_work> rocket, Right we discussed that previously I think.

[13-Sep-2011 10:43:37] <photon01> hi every1, I could use some help changing datapoints

[13-Sep-2011 10:43:41] <photon01> any1 can hel

[13-Sep-2011 10:43:43] <photon01> help

[13-Sep-2011 10:54:35] <konvert> anyone good at event transforms

[13-Sep-2011 11:04:28] <photon01> i did about 2 transforms

[13-Sep-2011 11:04:32] <photon01> maybe I can help

[13-Sep-2011 11:04:58] <konvert> i'm trying to transform the perf/memory

[13-Sep-2011 11:05:36] <konvert> like in this post docs/DOC-8346

[13-Sep-2011 11:05:49] <konvert> not the post itself, but in the comments

[13-Sep-2011 11:06:27] <konvert> get this error :

[13-Sep-2011 11:06:30] <konvert> Problem with line 23: p = (usedBytes / totalBytes) * 100

[13-Sep-2011 11:06:41] <konvert> exception           ZeroDivisionError: float division

[13-Sep-2011 11:07:46] <wayne___> sounds like totalBytes is ending up 0

[13-Sep-2011 11:07:51] <photon01> yes

[13-Sep-2011 11:07:56] <photon01> more like a python error

[13-Sep-2011 11:08:09] <photon01> maybe u didn"t actually get the value of totalBytes

[13-Sep-2011 11:09:50] <Hackman238> konvert: You'll want to be sure the variables are not None and >0

[13-Sep-2011 11:10:04] <Hackman238> konvert: Use if None not in [var1, var2]:

[13-Sep-2011 11:10:49] <konvert> phew that's way beyond my python programming skills

[13-Sep-2011 11:11:03] <Hackman238> konvert: Don't worry you'll learn

[13-Sep-2011 11:11:08] <photon01> try changing the summary to : evt.summary="total byte ="+totalBytes

[13-Sep-2011 11:11:08] <konvert>

[13-Sep-2011 11:11:18] <photon01> to check if that s your problem

[13-Sep-2011 11:11:38] <konvert> i think it's the way zenoss reads the memory of the server,

[13-Sep-2011 11:12:10] <konvert> think this transform transforms the memory when read from wmi, and i read it from snmp, can that be the problem?

[13-Sep-2011 11:14:10] <photon01> I didnt understand that sentence

[13-Sep-2011 11:14:22] <konvert> woops

[13-Sep-2011 11:15:36] <konvert> so i'm using snmp to read my memory numbers, and other people use wmi, and i think the code for the transform i try to use takes the variables for wmi and not them from snmp

[13-Sep-2011 11:15:46] <Hackman238> konvert: Can you post the whole transform?

[13-Sep-2011 11:15:55] <konvert> english is not my native language

[13-Sep-2011 11:15:57] <konvert> yes i can

[13-Sep-2011 11:16:13] <konvert> import re

[13-Sep-2011 11:16:13] <konvert> # Extract the percent and free from the summary

[13-Sep-2011 11:16:13] <konvert> m = re.search("threshold of [^:]+: current value ([\d\.]+)", evt.summary)

[13-Sep-2011 11:16:13] <konvert> if m:

[13-Sep-2011 11:16:13] <konvert>     from Products.ZenUtils.Utils import convToUnits

[13-Sep-2011 11:16:14] <konvert>     isWindows = re.search("PagingFileTotalUsage", evt.eventKey)

[13-Sep-2011 11:16:14] <konvert>     if isWindows:

[13-Sep-2011 11:16:15] <konvert>         # windows displays percentage used

[13-Sep-2011 11:16:15] <konvert>         p = float(m.groups()[0])

[13-Sep-2011 11:16:16] <konvert>         totalBytes = device.os.totalSwap

[13-Sep-2011 11:16:16] <konvert>         usedBytes = totalBytes * (p / 100)

[13-Sep-2011 11:16:17] <konvert>         freeBytes = totalBytes - usedBytes

[13-Sep-2011 11:16:17] <konvert>     else:

[13-Sep-2011 11:16:18] <konvert>         freeBytes = float(m.groups()[0]) * 1024

[13-Sep-2011 11:16:56] <konvert> flood protection

[13-Sep-2011 11:17:22] <konvert> you've got everything?

[13-Sep-2011 11:17:44] <SEJeff_work> konvert, Paste to somewhere like dpaste.org just to make sure we got it all please

[13-Sep-2011 11:19:49] <konvert> it is in here

[13-Sep-2011 11:19:51] <konvert> docs/DOC-8346

[13-Sep-2011 11:19:56] <konvert> the post of Dave Ta

[13-Sep-2011 11:20:12] <konvert> the second comment

[13-Sep-2011 11:33:55] <Hackman238> konvert: Should work

[13-Sep-2011 11:35:03] <konvert> i got error : ZeroDivisionError: float division -           Problem with line 23: p = (usedBytes / totalBytes) * 100

[13-Sep-2011 11:36:59] <dhopp> why am I failing to find the docs on putting zenoss behind apache

[13-Sep-2011 11:43:05] <SEJeff_work> dhopp, Because your google-foo is weak?

[13-Sep-2011 11:43:34] <SEJeff_work> dhopp, http://tinyurl.com/6elc3yz

[13-Sep-2011 11:43:38] <dhopp> SEJeff_work:  indeed it was…I found it

[13-Sep-2011 11:49:20] <photon01> how can I change like 50 datapoints easily

[13-Sep-2011 11:49:26] <photon01> maybe with zendmd ?

[13-Sep-2011 12:00:01] [disconnected at Tue Sep 13 12:00:01 2011]

[13-Sep-2011 12:00:02] [connected at Tue Sep 13 12:00:02 2011]

[13-Sep-2011 12:00:15] [zenoss-logger (logger bot) has joined #zenoss]

[13-Sep-2011 12:05:07] <photon01> any1 here ?

[13-Sep-2011 12:05:17] <Hackman238> photon01: Hey, whats up?

[13-Sep-2011 12:05:35] <Hackman238> konvert: I'm not sure why you're getting an error, the transform looks good

[13-Sep-2011 12:06:06] <photon01> <Hackman238> : Hi, I'm looking for a way to automatically alter 50datapoints

[13-Sep-2011 12:06:14] <photon01> <Hackman238> : any idea ?

[13-Sep-2011 12:06:31] <Hackman238> photon01: It depends on what you're trying to alter.

[13-Sep-2011 12:06:52] <photon01> <Hackman238> : rrd type and max value

[13-Sep-2011 12:07:40] <Hackman238> photon01: Okay. If you do that you'll need to delete your rrd's

[13-Sep-2011 12:07:58] <photon01> <Hackman238> : Yes, no problem with that

[13-Sep-2011 12:12:31] <photon01> <Hackman238> : so, any idea ?

[13-Sep-2011 12:13:34] <Hackman238> photon01: Trying to figure it out for you

[13-Sep-2011 12:14:42] <photon01> <Hackman238> : thanks ^^, I'm thinking about zendmd, but I dont know where to find the datapoints to modify them

[13-Sep-2011 12:14:48] <Hackman238> photon01: d=dmd.Devices.rrdTemplates._getOb('ZenJMX')

[13-Sep-2011 12:15:08] <Hackman238> photon01: this will make 'd' the rrdTemplate object of your choice

[13-Sep-2011 12:15:24] <Hackman238> photon01: Then in d.datasources you can change the attributes you need

[13-Sep-2011 12:15:56] <photon01> <Hackman238> : okaay, let me check

[13-Sep-2011 12:16:27] <photon01> <Hackman238> : by rrd template you mean monitoring template ?

[13-Sep-2011 12:16:42] <Hackman238> photon01: To select a datasource, i=d.datasources._getOb('ZenJMX Heap Memory')

[13-Sep-2011 12:16:44] <Hackman238> photon01: Yep

[13-Sep-2011 12:16:50] kocolosk_ is now known as kocolosk

[13-Sep-2011 12:16:57] <Hackman238> photon01: So i will be the datasource onject

[13-Sep-2011 12:18:05] <photon01> >>> d=dmd.Devices.rrdTemplates._getObj()

[13-Sep-2011 12:18:05] <photon01> Traceback (most recent call last):

[13-Sep-2011 12:18:05] <photon01>   File "<console>", line 1, in <module>

[13-Sep-2011 12:18:05] <photon01> AttributeError: _getObj

[13-Sep-2011 12:18:22] <Hackman238> _getOb

[13-Sep-2011 12:18:25] <photon01> oops

[13-Sep-2011 12:18:26] <Hackman238> not _getObj

[13-Sep-2011 12:19:13] <Hackman238> photon01: next, j=i.datapoints._getOb('ThreadCount')

[13-Sep-2011 12:19:30] <Hackman238> photon01: now j is the datapoint object

[13-Sep-2011 12:19:53] <Hackman238> photon01: set j.rrdmax and j.rrdtype

[13-Sep-2011 12:20:01] <Hackman238> photon01:

[13-Sep-2011 12:20:43] <photon01> nice ^^

[13-Sep-2011 12:20:45] <photon01> well done !

[13-Sep-2011 12:20:53] <Hackman238> photon01: Anytime

[13-Sep-2011 12:21:03] <photon01> so all i need is a python script that zendmd would execute

[13-Sep-2011 12:21:24] <Hackman238> photon01: Yep.

[13-Sep-2011 12:21:31] <photon01> <Hackman238> : another question plz

[13-Sep-2011 12:21:44] <photon01> <Hackman238> : unrelated, though

[13-Sep-2011 12:21:54] <Hackman238> photon01: Sure

[13-Sep-2011 12:22:11] <photon01> <Hackman> : any idea as to how can I have multiple dashboards for the same user ?

[13-Sep-2011 12:22:27] <Hackman238> photon01: Ah- yes and no.

[13-Sep-2011 12:22:39] <photon01> <Hackman> : I'll take the yes

[13-Sep-2011 12:22:54] <Hackman238> photon01: Its doable, but I dont have a solution implemented.

[13-Sep-2011 12:23:16] <photon01> <Hackman> : okay, is it complicated though ?

[13-Sep-2011 12:23:20] <Hackman238> photon01: in the ZMI under a user -> properties tab you'll find 'dashboardstate' or the like

[13-Sep-2011 12:23:36] <photon01> yes

[13-Sep-2011 12:23:37] <Hackman238> photon01: It has clear text in it. This text defines the dashboard.

[13-Sep-2011 12:23:50] <photon01> <Hackman> : i saw it, it has all the urls I set on the dashboard

[13-Sep-2011 12:24:18] <Hackman238> photon01: What I've done in the past is write a script that updates that string to be the dashboard I need to use at a certain time

[13-Sep-2011 12:24:53] <photon01> <Hackman> : I see, using zendmd ?

[13-Sep-2011 12:25:00] <Hackman238> photon01: Unfortunately its not a good solution. Generally speaking, I'd tell users to pick a dash and stick with it.

[13-Sep-2011 12:25:21] <Hackman238> photon01: Yeah use a python script, import ZenScriptBase and set a dmd object

[13-Sep-2011 12:25:40] <photon01> <Hackman> : okaay

[13-Sep-2011 12:25:49] <photon01> <Hackman> :  so I guess there's no easy way for now

[13-Sep-2011 12:25:57] <Hackman238> photon01: You could even create a little portal that executes teh script and stores alternative dashboard in the users properties under a new variable.

[13-Sep-2011 12:26:11] <Hackman238> photon01: Unfortunately not.

[13-Sep-2011 12:26:41] <photon01> <Hackman> : would have been nice to have tabs for multiple dashboards

[13-Sep-2011 12:27:03] <photon01> <Hackman> : I think I'm going to post something on forum for that

[13-Sep-2011 12:27:11] <Hackman238> photon01: yeah, its a great idea.

[13-Sep-2011 12:27:16] <photon01> <Hackman> : thanks again

[13-Sep-2011 12:27:38] <Hackman238> photon01: Do me a favor, email hackman238@gmail.com a synopsis and I'll add it to the ZCA devel roadmap which will be out later today

[13-Sep-2011 12:27:49] <Hackman238> photon01: NP

[13-Sep-2011 12:28:56] <photon01> <Hackman> : okaay ^^

[13-Sep-2011 12:46:19] <dhopp> Hackman238:  what options do you use in apache to load balance zopes?

[13-Sep-2011 12:53:55] <Hackman238> dhopp: I'll email you a template

[13-Sep-2011 13:10:37] <Hackman238> dhopp: Emailed

[13-Sep-2011 13:28:54] <Jane_Curry> hiya

[13-Sep-2011 13:29:16] <Jane_Curry> Has anyone  seen any traps in a Zenoss 3.2 processed by zentrap??????

[13-Sep-2011 13:32:43] <Hackman238> Jane_Curry: I haven't tried. Zentrap not working?

[13-Sep-2011 13:36:20] <Jane_Curry> I am generating them -  I can see them in the raw incoming file - but nothing ever gets processed by zentrap

[13-Sep-2011 13:36:51] <hmp> nothing in logs?

[13-Sep-2011 13:37:59] <Hackman238> Jane_Curry: Interesting. Anything from zenhub in debug mode?

[13-Sep-2011 13:38:35] <Jane_Curry> Not obviously - I now have everything locked up! zenoss stop | start time

[13-Sep-2011 13:38:51] <dhopp> Hackman238:  thanks…looking at it now

[13-Sep-2011 13:39:28] <dhopp> Hackman238:  holy crap..how many zopes do you need running?

[13-Sep-2011 13:48:42] <Hackman238> dhopp: 40 on the small boxes

[13-Sep-2011 13:48:49] <dhopp> on the "small" boxes?

[13-Sep-2011 13:48:50] <Hackman238> dhopp: 80 on the major boxes

[13-Sep-2011 13:49:00] <Hackman238> dhopp: Yep

[13-Sep-2011 13:49:11] <dhopp> how many users are accessing "small"?

[13-Sep-2011 13:49:20] <Hackman238> dhopp: 256GB RAM in those puppies

[13-Sep-2011 13:49:26] <Hackman238> dhopp: Many many

[13-Sep-2011 13:49:46] <dhopp> I was thinking of 4 or 5 zopes for ~30 concurrent users (peak time would be more)

[13-Sep-2011 13:49:52] <Hackman238> dhopp: Also users often use the 'show all' on a special page we created that shows all interface graphs for a device on a tab

[13-Sep-2011 13:50:19] <Hackman238> dhopp: Which can result in hundreds of graphs on big devices, and nearly a thousand on major cores

[13-Sep-2011 13:50:46] <dhopp> Hackman238:  thousand graphs on a page…that's not hard to go through and figure out what you want to see

[13-Sep-2011 13:50:47] <dhopp> lol

[13-Sep-2011 13:51:19] <Hackman238> dhopp: Their serialized nice and neat. Its so the NOC folks can look at every int very quickly to find DDoS and other problems

[13-Sep-2011 13:51:30] <Hackman238> *s/Their/They're/

[13-Sep-2011 13:53:22] <dhopp> Hackman238:  I'm assuming you patched zope as well to support the multiple instances?

[13-Sep-2011 13:53:49] <Hackman238> dhopp: Negative. I've a multizope pack I developed. It's planned for release to the community

[13-Sep-2011 13:53:57] <dhopp> oh!

[13-Sep-2011 13:54:01] <Hackman238> dhopp: Just modifications to the zope control scripts

[13-Sep-2011 13:54:15] <Hackman238> dhopp: Tonight I'll release the ZCA devel roadmap

[13-Sep-2011 13:54:33] <Hackman238> dhopp: I meant to yesterday but I've felt like crap since my nuts hike this past sunday

[13-Sep-2011 13:54:51] <dhopp> that would be handy to have…I'll look forward to that but for now I'll use the multizope patch that is available on community.zenoss.org

[13-Sep-2011 13:55:07] <dhopp> "nuts hike"…I won't ask

[13-Sep-2011 13:56:11] <Hackman238> dhopp: 9mi on a rocky, uphill trail here in Texas, 9mi back. Ran out of water at 45% point LOL

[13-Sep-2011 13:57:14] <Jane_Curry> The joys of 3.2!!!  It seems to have completely screwed both Zope and the Events database - never see such a well-hosed system!

[13-Sep-2011 13:58:39] <Hackman238> Jane_Curry: Yikes. Whats happening?

[13-Sep-2011 14:05:58] <JohnnyNOC> :[

[13-Sep-2011 14:07:37] <Hackman238> JohnnyNOC: Whats up?

[13-Sep-2011 14:08:13] <JohnnyNOC> not too much

[13-Sep-2011 14:08:18] <JohnnyNOC> just ya know, work..  lots of it.

[13-Sep-2011 14:08:34] <JohnnyNOC> wanting to upgrade to 3.2 but seems to be some trouble afoot

[13-Sep-2011 14:09:38] <hmp> ill wait for 4.0

[13-Sep-2011 14:09:58] <hmp> doesnt seem worth it, few benefites for a lot of potential problems

[13-Sep-2011 14:20:29] <JohnnyNOC> yea, unfortunately thats what it seems like..  but how long til 4.0?  :/

[13-Sep-2011 14:21:17] <hmp> i saw a roadmap saying december

[13-Sep-2011 14:27:52] <Simon4> there must be a bunch-o-people on 2.5.x who are going to have a hell of an upgrade path to 4 when it hits and they have to jump through the versions

[13-Sep-2011 14:32:45] <JohnnyNOC> geez

[13-Sep-2011 14:33:07] <JohnnyNOC> hope that's not the case..  just spent a lot of time with a fresh 3.1.0 install

[13-Sep-2011 14:33:34] <Simon4> JohnnyNOC: my guess is 2.5->3.something->4 is the path

[13-Sep-2011 14:34:00] <JohnnyNOC> i don't seem to have many problems with 3.1.0

[13-Sep-2011 14:36:24] <JohnnyNOC> learning to like it in fact

[13-Sep-2011 14:36:55] <Hackman238> JohnnyNOC: You must upgrade to v3 to upgrade to 4

[13-Sep-2011 14:37:14] <Hackman238> JohnnyNOC: Unofficially, 4.0 for Core should be out in a few months

[13-Sep-2011 14:43:16] <JohnnyNOC> Hackman238 will you hold my hand through the upgrade

[13-Sep-2011 14:44:17] <Jane_Curry> Hackman238:  Dunno - no events; no devices and event.log is getting periodic

[13-Sep-2011 14:44:40] <Jane_Curry> 2011-09-13T19:43:04 ERROR ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 127.0.0.1:57215 at 0xab14b4c channel#: 75 requests:> (<type 'exceptions.ValueError'>:HTTP headers invalid (too long) (got: 8227 bytes, allowed 8192 bytes [/usr/local/zenoss/python/lib/python2.6/asyncore.py|read|74] [/usr/local/zenoss/python/lib/python2.6/asyncore.py|handle_read_event|413] [/usr/local/zenoss/python/

[13-Sep-2011 14:44:41] <Jane_Curry> lib/python2.6/asynchat.py|handle_read|155] [/usr/local/zenoss/python/lib/python2.6/site-packages/Zope2-2.12.1-py2.6-linux-i686.egg/ZServer/HTTPServer.py|collect_incoming_data|435])

[13-Sep-2011 14:45:39] <Jane_Curry> Navigating to INFRASTRUCTURE -> Devices jsut gets a permanent "Loading Please wait" message

[13-Sep-2011 14:50:57] <Hackman238> JohnnyNOC: Not today, but sure

[13-Sep-2011 14:51:13] <Hackman238> Jane_Curry: Wow...thats a serious error

[13-Sep-2011 14:52:23] <Hackman238> Jane_Curry: Lets tweak your zope.conf

[13-Sep-2011 14:53:19] <Hackman238> Jane_Curry: In your zope.conf uncomment 'http-header-max-length 16384' and restart zope

[13-Sep-2011 14:53:21] <Jane_Curry> suggestions????

[13-Sep-2011 14:53:32] <Hackman238> Jane_Curry: Let me know if that fixes the problem

[13-Sep-2011 14:53:42] <Hackman238> JohnnyNOC: When do you plan on migrating?

[13-Sep-2011 14:53:53] <Jane_Curry> I've tried zenfixit - which didn't barf - contrary to a forum append recently - but it didn't fix things

[13-Sep-2011 14:54:42] <Hackman238> Jane_Curry: Let me know if that zope.conf directive changes anything

[13-Sep-2011 14:58:44] <Jane_Curry> Hackman238 - thou workest good magic! I have devices and no more errors in event.log

[13-Sep-2011 14:59:04] <Jane_Curry> Question is - how did I get to that state and why will I not get there again???

[13-Sep-2011 14:59:07] <Hackman238> Jane_Curry: Glad its working

[13-Sep-2011 14:59:26] <Hackman238> Jane_Curry: That directive limits the len to prevent DoS attack

[13-Sep-2011 14:59:43] <Hackman238> Jane_Curry: You're values are falling just outside it for some reason...IDK why

[13-Sep-2011 14:59:54] <Hackman238> Jane_Curry: I've had this problem randomly. No idea why it happens

[13-Sep-2011 15:00:42] <Hackman238> Jane_Curry: Anywho...your zope is more suceptible to attack...but your Zenoss shouldn't be web accessible anyways

[13-Sep-2011 15:01:44] <Hackman238> Ugh...s/you're/your/

[13-Sep-2011 15:02:08] <Hackman238> I have to disable this damn plugin. LOL its killing me

[13-Sep-2011 15:05:06] <Jane_Curry> Still don't get any traps processed by zentrap though

[13-Sep-2011 15:05:50] <Jane_Curry> Even though I have raw trap logging turned on and the box is definitely receiving traps

[13-Sep-2011 15:06:11] <Jane_Curry> Can anyone else confirm / deny any trap reception with 3.2??

[13-Sep-2011 15:06:45] <bender183> we still use 2.4.2

[13-Sep-2011 15:06:46] <bender183> =\

[13-Sep-2011 15:07:47] <dhopp2> Jane_Curry: clear your browser cache

[13-Sep-2011 15:09:10] <dhopp2> sorry..just catching up..didn't see that Hackman238 had already solved it

[13-Sep-2011 15:09:30] dhopp2 is now known as dhopp

[13-Sep-2011 15:10:25] <Jane_Curry> I also have an append on zenoss-users at message/61327#61327...

[13-Sep-2011 15:10:40] <Jane_Curry> All my processes are hosed as far as seeing them goes

[13-Sep-2011 15:11:16] <Jane_Curry> They do work but you cannot inspect them or change their parameters - just get a "Loading" message

[13-Sep-2011 15:11:41] <Jane_Curry> Bah - there has been a lot of gratuitous "loading" today!

[13-Sep-2011 15:11:47] <dhopp> Jane_Curry:  what browswer?

[13-Sep-2011 15:11:55] <dhopp> *browser

[13-Sep-2011 15:12:11] <Jane_Curry> Firefox 3.6.10

[13-Sep-2011 15:12:46] <Hackman238> Jane_Curry: I dont understand

[13-Sep-2011 15:12:48] <dhopp> hrm…I saw something similar with chrome and I cleared chomes cookies and the problem went away

[13-Sep-2011 15:13:41] <Hackman238> Jane_Curry: Oh I see. Looks like a router (in python) problem.

[13-Sep-2011 15:14:13] <Hackman238> Jane_Curry: So zenchkrels -x1 -r barfs?

[13-Sep-2011 15:14:26] <Hackman238> Jane_Curry: Try to avoid zenfixit...it can unfix a lot

[13-Sep-2011 15:14:50] <Hackman238> Jane_Curry: You can also try method 2 here http://shanewilliamscott.com/Code/zendmd-tip-fix-broken-relationships.html

[13-Sep-2011 15:16:00] <dhopp> Hackman238:  woo…so I setup a basic apache proxy infront of zenoss and since doing that chrome would give me an error loading the page but firefox worked fine….I just changed that http-header-max-length variable and restarted zope and now chrome works

[13-Sep-2011 15:16:07] <dhopp> if I change it back..chrome breaks

[13-Sep-2011 15:16:39] <Hackman238> dhopp: Just leave the directive uncommented

[13-Sep-2011 15:16:51] <Hackman238> dhopp: Zenoss should never be exposed to anyone who will DoS it anyway

[13-Sep-2011 15:17:01] <Jane_Curry> Hmmm - just tried with Firefox 4.0.b12 and processes are OK

[13-Sep-2011 15:17:40] <Hackman238> Jane_Curry: Should work with FF4,5,6

[13-Sep-2011 15:17:45] <dhopp> Hackman238:  I'm fine leaving it uncommented..but makes me wonder what chrome is doing to require headers

[13-Sep-2011 15:17:47] <dhopp> that big

[13-Sep-2011 15:17:56] <Hackman238> Jane_Curry: I know some UI changes may have broken the AJAX compatability with 3.6.X

[13-Sep-2011 15:18:13] <Hackman238> dhopp: Who knows....thats why I wont support chrome.

[13-Sep-2011 15:18:25] <Hackman238> dhopp: FF, Opera or go elsewhere

[13-Sep-2011 15:18:27] <Hackman238> LOL

[13-Sep-2011 15:20:17] <dhopp> Hackman238:  not even IE ;-)

[13-Sep-2011 15:20:32] <Jane_Curry> dhopp - you're right I cleared cache and that didn't help.  I cleared cookies and I now have my processes back of FF 3.6.10 - thanks

[13-Sep-2011 15:21:45] <dhopp> Hackman238:  if zenfixit 'unfixes' things..why do people seem to always recommend it?  (I've never personally used it)...

[13-Sep-2011 15:22:23] <Hackman238> dhopp: Not officially, no.

[13-Sep-2011 15:22:48] <Hackman238> dhopp: We can only support so much. If people want to be a one off they can futz with their box all they want to make it work.

[13-Sep-2011 15:23:04] <Hackman238> Jane_Curry: Interesting....corrupt cookie

[13-Sep-2011 15:23:13] <Jane_Curry> Still no traps though

[13-Sep-2011 15:23:20] <Hackman238> Jane_Curry: I ran in to that same problem the other day with a 4.0.2 test

[13-Sep-2011 15:23:54] <Hackman238> Jane_Curry: Can you 'zenhub debug' and 'zentrap debug' and recheck the logs?

[13-Sep-2011 15:24:06] <Hackman238> Jane_Curry: Less they are already in debug

[13-Sep-2011 15:24:10] <Jane_Curry> I have zentrap logging turned right up and there is an entry for capturing packet... and Received packet.... but nothing else - no processing

[13-Sep-2011 15:24:29] <Hackman238> dhopp: Zenfix i is dangerous...it can cause problems in some cases. Try to avoid it except in last ditch efforts

[13-Sep-2011 15:24:43] <Hackman238> Jane_Curry: Wow. Thats odd.

[13-Sep-2011 15:33:50] <Jane_Curry> Yeh - turned up logging on zenhub but don't see anything good

[13-Sep-2011 15:33:58] <Jane_Curry> Found the following in zentrap

[13-Sep-2011 15:34:16] <Jane_Curry> 2011-09-13 20:25:47,946 INFO zen.zentrap: Starting to listen on SNMP trap port 162

[13-Sep-2011 15:34:17] <Jane_Curry> 2011-09-13 20:25:47,946 Level 21 zen.netsnmp: registered debug token snmp_parse, 1

[13-Sep-2011 15:34:19] <Jane_Curry> 2011-09-13 20:25:47,947 DEBUG zen.netsnmp: Cannot find constructor function for UDP/IPv6 transport domain object.

[13-Sep-2011 15:35:38] <Hackman238> Jane_Curry: That just means IPV6 isnt enabled on that box

[13-Sep-2011 15:35:55] <Hackman238> Jane_Curry: iptables is clean?

[13-Sep-2011 15:39:33] <Jane_Curry> If I am seeing the raw traps coming in then I doubt it's firewall stuff - these are traps from the zenoss box itself

[13-Sep-2011 15:39:52] <Hackman238> Jane_Curry: Oh right. Doh

[13-Sep-2011 15:41:19] <Hackman238> Jane_Curry: Does the box have multiple nics?

[13-Sep-2011 15:41:45] <Hackman238> Jane_Curry: and is snmptrapd off?

[13-Sep-2011 15:42:32] <Jane_Curry> single NIC - it uses it's own real address - not 127.0.0.2 / localhost

[13-Sep-2011 15:43:03] <Jane_Curry> zentrap is bound to UDP *:162 like it should be

[13-Sep-2011 15:52:27] <Hackman238> Jane_Curry: so zentrap is set ti listenip {real ip}?

[13-Sep-2011 15:53:42] <Jane_Curry> Sorry Hackman238 - don't understand the last comment.....

[13-Sep-2011 15:55:55] <Jane_Curry> My lsof on my 3.2 system looks exactly comparable with my 3.1 system

[13-Sep-2011 15:57:32] <Hackman238> Jane_Curry: Sorry, in $zenhome/etc/zentrap.conf, is listenip set to the real local ip?

[13-Sep-2011 15:57:37] <Hackman238> Jane_Curry: Gotcha

[13-Sep-2011 15:59:13] <Jane_Curry> Hmmm - listenip - 0.0.0.0 - should I be worried?

[13-Sep-2011 16:01:21] <Hackman238> Jane_Curry: No, 0.0.0.0 means it binds to everything.

[13-Sep-2011 16:01:48] <Hackman238> Jane_Curry: Lets try setting it to the ip and restarting the daemon any way.

[13-Sep-2011 16:01:59] <Jane_Curry> 0.0.0.0 seems to be the default and I expect that to mean bind to everything

[13-Sep-2011 16:02:15] <Hackman238> Jane_Curry: Thats correct

[13-Sep-2011 16:05:46] <Jane_Curry> Tried hard coding listenip and recycled zentrap - not a sausage-worth of difference

[13-Sep-2011 16:07:33] <Hackman238> Jane_Curry: Darn.

[13-Sep-2011 16:07:47] <Hackman238> Jane_Curry: I'll be honest Jane, I'm out of ideas

[13-Sep-2011 16:08:02] <Hackman238> Jane_Curry: You might want to hit up rocket or klone

[13-Sep-2011 16:14:55] <Jane_Curry> I've put on a forum append - don't know how to bug developers directly unless they are in IRC

[13-Sep-2011 16:15:23] <Jane_Curry> The "£*&$%£" have also broken the mib_browser ZenPack AGAIN!

[13-Sep-2011 16:15:47] <Hackman238> Jane_Curry: Klone's going to eyeball the problem right now

[13-Sep-2011 16:15:58] <psyklone> one sec, Jane_Curry

[13-Sep-2011 16:17:23] <Hackman238> BRB

[13-Sep-2011 16:25:03] <Hackman238> Back

[13-Sep-2011 16:28:38] <Jane_Curry> Thanks psyklone

[13-Sep-2011 16:33:00] <psyklone> Jane_Curry, can you run 'zentrap debug' to pop it into debug mode and then tail $ZENHOME/log/localhost/zentrap.log as you fire off a few traps?

[13-Sep-2011 16:35:03] <Jane_Curry> I have zentrap set to debug..

[13-Sep-2011 16:35:30] <Jane_Curry> and I have capture raw packets set to on

[13-Sep-2011 16:35:46] <Jane_Curry> I can see files being created for the raw input when traps arrive

[13-Sep-2011 16:36:11] <Jane_Curry> I've tried authentication, cold start, warm start and an snmptrap test script that I have

[13-Sep-2011 16:36:20] <Jane_Curry> they all generate raw input files

[13-Sep-2011 16:36:29] <Jane_Curry> but zentrap.log only shows

[13-Sep-2011 16:38:46] <Jane_Curry> 2011-09-13 21:37:01,547 DEBUG zen.thresholds: Checking value 0 on Daemons/localhost/zentrap_eventQueueLength

[13-Sep-2011 16:38:47] <Jane_Curry> 2011-09-13 21:37:01,547 DEBUG zen.MinMaxCheck: Checking zentrap_eventQueueLength 0 against min None and max 1000

[13-Sep-2011 16:38:49] <Jane_Curry> 2011-09-13 21:37:07,296 DEBUG zen.zentrap: Received packet from 192.168.10.131 at port 23169

[13-Sep-2011 16:38:50] <Jane_Curry> 2011-09-13 21:37:07,296 DEBUG zen.zentrap: Capturing packet from 192.168.10.131

[13-Sep-2011 16:38:52] <Jane_Curry> 2011-09-13 21:37:27,554 DEBUG zen.zentrap: Received packet from 192.168.10.131 at port 5249

[13-Sep-2011 16:38:53] <Jane_Curry> 2011-09-13 21:37:27,554 DEBUG zen.zentrap: Capturing packet from 192.168.10.131

[13-Sep-2011 16:38:55] <Jane_Curry> 2011-09-13 21:37:28,566 DEBUG zen.zentrap: Received packet from 192.168.10.131 at port 5249

[13-Sep-2011 16:38:56] <Jane_Curry> 2011-09-13 21:37:28,567 DEBUG zen.zentrap: Capturing packet from 192.168.10.131

[13-Sep-2011 16:38:58] <Jane_Curry> 2011-09-13 21:37:29,569 DEBUG zen.zentrap: Received packet from 192.168.10.131 at port 5249

[13-Sep-2011 16:38:59] <Jane_Curry> 2011-09-13 21:37:29,570 DEBUG zen.zentrap: Capturing packet from 192.168.10.131

[13-Sep-2011 16:39:01] <Jane_Curry> 2011-09-13 21:37:30,574 DEBUG zen.zentrap: Received packet from 192.168.10.131 at port 5249

[13-Sep-2011 16:39:02] <Jane_Curry> 2011-09-13 21:37:30,575 DEBUG zen.zentrap: Capturing packet from 192.168.10.131

[13-Sep-2011 16:39:04] <Jane_Curry> 2011-09-13 21:37:31,576 DEBUG zen.zentrap: Received packet from 192.168.10.131 at port 5249

[13-Sep-2011 16:39:05] <Jane_Curry> 2011-09-13 21:37:31,576 DEBUG zen.zentrap: Capturing packet from 192.168.10.131

[13-Sep-2011 16:39:07] <Jane_Curry> 2011-09-13 21:37:32,584 DEBUG zen.zentrap: Received packet from 192.168.10.131 at port 5249

[13-Sep-2011 16:39:08] <Jane_Curry> 2011-09-13 21:37:32,585 DEBUG zen.zentrap: Capturing packet from 192.168.10.131

[13-Sep-2011 16:39:17] <Jane_Curry> ... thats an snmptrap script plus an snmpwalk with a bad community name

[13-Sep-2011 16:39:27] <Jane_Curry> There's just no processing going on

[13-Sep-2011 16:46:45] <psyklone> Jane_Curry - sorry, was looking at the results i'm getting here ... which are a bit different. i take it you're sending v2 traps?

[13-Sep-2011 16:49:29] <Jane_Curry> Nope - v1 traps

[13-Sep-2011 16:50:28] <psyklone> and this is on 3.2?

[13-Sep-2011 16:51:54] <Jane_Curry> yup 3.2 core - Open SuSE 32-bit stack build

[13-Sep-2011 16:52:46] <Hackman238> Jane_Curry: Why does your log say 'at port 5249'?

[13-Sep-2011 16:59:17] <Jane_Curry> presumably that is the sending, ephemeral port - those are the authentication failure traps

[13-Sep-2011 17:15:33] <psyklone> Jane_Curry, i think i've got a fix for you

[13-Sep-2011 17:27:42] <psyklone> yep, there's a trap issue

[13-Sep-2011 17:29:32] <psyklone> Jane_Curry, i've got a patch for you to load up that's fixed it here on my system. ping me when you get a chance

[13-Sep-2011 17:30:10] <Jane_Curry> ok

[13-Sep-2011 17:45:41] <Jane_Curry> Looks like that patch fixed it - when willl it be official and do we have a number for it??

[13-Sep-2011 17:46:35] <psyklone> thanks for the verification ... i've created an internal defect on this and attached the patch. it'll get merged in for the maintenance release, and in the meantime will be zenpatchable in the very near future

[13-Sep-2011 18:47:15] <tvincent> What may be a silly question: But I am looking at the http://dev.zenoss.org/trac/browser/trunk/zenpacks/ZenPacks.zenoss.MySqlMonitor/ZenPacks/zenoss/MySqlMonitor zenpack. Where are the queries to get the actually stats?

[13-Sep-2011 19:25:18] kocolosk is now known as kocolosk|away

[13-Sep-2011 20:31:12] <Tabmow> Anyone monitoring postgresql in zenoss? I dded the community packs but having a hard time setting it up and binding the templates

[14-Sep-2011 00:00:01] [disconnected at Wed Sep 14 00:00:01 2011]

[14-Sep-2011 00:00:02] [connected at Wed Sep 14 00:00:02 2011]

[14-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[14-Sep-2011 03:18:50] froztbyt1 is now known as froztbyte

[14-Sep-2011 07:35:21] <hmp> is it possible to put a collecor in some kind of maintenence mode

[14-Sep-2011 07:35:31] <hmp> so no hb errors come up

[14-Sep-2011 08:18:26] klone is now known as psyklone

[14-Sep-2011 08:25:59] jb____ is now known as jb

[14-Sep-2011 08:42:41] <Hackman238> Morning all!

[14-Sep-2011 08:42:46] <psyklone> howdy

[14-Sep-2011 09:06:14] <Hackman238> What's the new patch?

[14-Sep-2011 09:08:00] <psyklone> `zenpatch 26337`fixes an issue jane found yesterday afternoon with zentrap not processing traps

[14-Sep-2011 09:11:51] Guest29337 is now known as Glace

[14-Sep-2011 09:12:11] <Hackman238> psyklone: Very cool.

[14-Sep-2011 09:12:18] <Hackman238> psyklone: I PM'd you

[14-Sep-2011 09:12:21] Glace is now known as Guest48560

[14-Sep-2011 09:12:29] Guest48560 is now known as Glacee

[14-Sep-2011 09:43:04] <dpetzel> ok feeling like a total newb.... Looking to finally upgrade my 3.0.3 install. I found in the docs, I need to jump to 3.1 first prior to going to 3.2 However, this is my first time ever doing an upgrade, and I dont see any upgrade instructions. The closest thing I found was the upgrade section of the 3.0 install guide. Im I really this blind, or are there some other upgrade instuctions? Or is it so stupid simple it doesnt need instructions? (IE just inst

[14-Sep-2011 09:43:04] <dpetzel> all the RPM and your done)

[14-Sep-2011 09:48:58] kocolosk|away is now known as kocolosk

[14-Sep-2011 09:53:12] <psyklone> dpetzel, it's just the steps here: docs/DOC-10160

[14-Sep-2011 09:55:36] <dpetzel> psyklone: Thanks, that about mimics whats in the 3.0 guide. I curious only becuse I of course already have the pre-reqs yum packages as I needed them for the initial install, so the doc isnt a true 3.03 --> 3.1 upgrade guide, all that aside, it looks easy enough

[14-Sep-2011 09:56:56] <psyklone> dpetzel, yeah ... that isn't specifically a 3.0-3.1 upgrade guide, but those steps apply. it's a very easy upgrade to get to 3.1 from 3.0.3

[14-Sep-2011 09:57:17] <dpetzel> psyklone: Cool, thanks for the info

[14-Sep-2011 10:08:06] <tsener> hello

[14-Sep-2011 10:09:22] <tsener> having trouble with polling some snmp info

[14-Sep-2011 10:13:55] <tsener> when I query via snmpwalk  i get returned value..  but in zenoss appears debug message OID bad

[14-Sep-2011 10:15:19] <dhopp> tsener:  what's the OID?

[14-Sep-2011 10:16:32] <tsener> .1.3.6.1.4.1.1916.1.32.1.2

[14-Sep-2011 10:16:48] <tsener> this is on an exttreme switch, should return CPU utilization in percentage

[14-Sep-2011 10:17:31] <tsener> I create a template and use this OID as  SNMP  data source

[14-Sep-2011 10:18:01] <dhopp> and what happens if you do a 'test agains a device' from the data source?

[14-Sep-2011 10:18:54] <DW-10297> Is the windows 2008 SNMP agent terrible or is it just me?

[14-Sep-2011 10:20:40] mattray1 is now known as mattray

[14-Sep-2011 10:32:21] <rmatte> DW-10297: nope, I see no difference in the SNMP on 2008 than on 2003

[14-Sep-2011 10:32:35] <rmatte> It's a million times more reliable than WMI

[14-Sep-2011 10:33:14] <tsener> dhopp: I can't

[14-Sep-2011 10:33:21] <tsener> since it's snmpv3

[14-Sep-2011 10:33:42] <tsener> its zenoss 2.5.2 and for some reason it does not know about the auth details that I've set to it

[14-Sep-2011 10:33:58] <tsener> other than that it colelcts the resto fo the info via SNMP using v3

[14-Sep-2011 10:34:42] <DW-10297> ah, it just keeps telling me that the connection to wmi has failed then it says it's back up then it says its failed then it says its back up and this keeps repeating over and over

[14-Sep-2011 10:34:45] <DW-10297> it's really annoying

[14-Sep-2011 10:35:06] <rmatte> DW-10297: WMI isn't SNMP

[14-Sep-2011 10:35:28] <rmatte> WMI is a microsoft specific protocol and it blows ass, which is why I only use it in a very targetted fashion

[14-Sep-2011 10:35:52] <rmatte> I try to use SNMP as much as possible for monitoring windows servers

[14-Sep-2011 10:37:16] <DW-10297> Any idea what the command for the wmi thing is so I can try it in the command line?

[14-Sep-2011 10:37:22] <DW-10297> for zenoss

[14-Sep-2011 10:37:29] <DW-10297> is it zenwin?

[14-Sep-2011 10:38:24] <rmatte> yeh, I think zenwin is the daemon that handles that

[14-Sep-2011 10:38:37] <rmatte> but I guarantee you that if it's dropping like that it's the server's fault

[14-Sep-2011 10:38:42] <rmatte> we deal with that sort of thing constantly

[14-Sep-2011 10:38:51] <rmatte> WMI just decides to stop working

[14-Sep-2011 10:39:04] <DW-10297> That doesn't really make sense the server is operating normally

[14-Sep-2011 10:39:12] <rmatte> WMI is buggy man

[14-Sep-2011 10:39:25] <rmatte> and it's buggier on 2008 than it was on 2003

[14-Sep-2011 10:39:45] <rmatte> especially on 2008 SP1, it's slightly better on SP2 but still has issues

[14-Sep-2011 10:40:21] <rmatte> which is why I try to avoid it whenever possible

[14-Sep-2011 10:40:55] <illsci> I just setup my Zenoss and the nodes I've entered all are alerting for http and smtp and I'm not sure why....   It's /Status/IpServices/  IP service smtp is down and http is down and I'm not sure why it's monitoring those out of the box

[14-Sep-2011 10:41:01] <illsci> where do I fix that....

[14-Sep-2011 10:41:17] <Hackman238> rmatte: Did you have a chance to daemonize your formula pack?

[14-Sep-2011 10:41:39] <rmatte> Hackman238: unfortunately not, had a lot of other stuff on the go lately

[14-Sep-2011 10:41:55] <rmatte> have a friend in hospital that I've been visiting almost every night, and work has been death lol

[14-Sep-2011 10:42:20] <Hackman238> rmatte: Sorry to hear about your friend.

[14-Sep-2011 10:42:30] <rmatte> he's doing better, thanks

[14-Sep-2011 10:42:42] <Hackman238> rmatte: Glad to hear.

[14-Sep-2011 10:43:13] <Hackman238> illsci: WMI in Windows 2008 is at best slow and dodgey

[14-Sep-2011 10:43:32] <rmatte> I had begun work on an SNMP modeler/template for monitoring windows services via SNMP...

[14-Sep-2011 10:43:50] <Hackman238> rmatte: That would be great

[14-Sep-2011 10:43:53] <rmatte> but then I found out that the existing services skin is directly tied in to whether or not wmi services monitoring is enabled in zProperties

[14-Sep-2011 10:44:01] <rmatte> so looks like I'll have to make a new skin from scratch

[14-Sep-2011 10:44:04] <illsci> Hackman238: I don't understand...

[14-Sep-2011 10:44:08] <illsci> i don't have any windows boxes

[14-Sep-2011 10:44:10] <DW-10297> Do you also have to change the OIDs? as this doesn't appear to be valid

[14-Sep-2011 10:44:12] <DW-10297> 1.3.6.1.4.1.9600.1.1.2.2.0

[14-Sep-2011 10:44:15] <DW-10297> in Windows 2008

[14-Sep-2011 10:44:39] <rmatte> DW-10297: what is that OID for?

[14-Sep-2011 10:45:27] <Hackman238> illsci: My bad. Wwhomever was aking about WMI- try to avoid it on Win2K8

[14-Sep-2011 10:45:45] <tsener> so far I gather windows services  via WMI

[14-Sep-2011 10:45:54] <tsener> can I do this via SNMP on Windows ?

[14-Sep-2011 10:45:56] <rmatte> DW-10297: there is a default SNMP template for CPU for windows devices, but it relies on having SNMP Informant installed on your windows servers (which is a waste of time)... I made the following ZenPacks that can monitor CPU/Memory via the standard windows SNMP agent: docs/DOC-3386 docs/DOC-3570

[14-Sep-2011 10:46:04] <Hackman238> tsener: Yes

[14-Sep-2011 10:46:17] <rmatte> tsener: you can, but the method isn't finished being developed yet

[14-Sep-2011 10:46:47] <DW-10297> ohh, why in the world would the default monitoring use a 3rd party agent

[14-Sep-2011 10:47:26] <rmatte> DW-10297: because it was easier from a template perspective to do it the way they did

[14-Sep-2011 10:47:43] <rmatte> DW-10297: my packs have code in them that does a bunch of work to reliably gather the values

[14-Sep-2011 10:48:03] <rmatte> DW-10297: since the way the values are presented is a bit wonky

[14-Sep-2011 10:55:28] <Scout__> has anyone seen problems with modeling after 3.2.0 upgrade?  all information is being put in snmp desciption but hardware and os information is staying very generic

[14-Sep-2011 10:55:36] <tsener> adding 0 to the end of the OID  I was polling  fixed it, no more "Bad OID" errors

[14-Sep-2011 10:56:09] <tsener> I guess it is supposed to contain more than one value

[14-Sep-2011 10:56:55] <psyklone> well, i think that the issue is that it _can_ contain more than one ... so adding the .0 index does the trick

[14-Sep-2011 10:57:24] <tsener> 10 psyklone

[14-Sep-2011 10:57:28] <tsener> *10x

[14-Sep-2011 11:09:06] <dhopp> rmatte:  I thought it only used the SNMP informant for disk IO monitoring

[14-Sep-2011 11:09:25] <rmatte> dhopp: no, the default performance template that comes with Zenoss for windows CPU via SNMP depends on informant

[14-Sep-2011 11:09:44] <rmatte> dhopp: also, the disk IO monitoring doesn't require informant, it uses the UCD MIB

[14-Sep-2011 11:31:11] <tsener> When  I graph percentage values, how would I use fixed Y axis  size

[14-Sep-2011 11:31:45] <rmatte> tsener: set the unit value for the graph to "percentage"

[14-Sep-2011 11:31:46] <tsener> I mean I want to see how  values scale to 100%, my current values are all below 50% so the graph  scales up to this value

[14-Sep-2011 11:31:49] <rmatte> and RRD will do it automatically

[14-Sep-2011 11:39:42] <rmatte> it's great, got it all rooted and have cyanogenmod (modded firmware) installed on it.

[14-Sep-2011 11:39:51] <rmatte> err, that was meant for privmsg obviously lol

[14-Sep-2011 12:00:01] [disconnected at Wed Sep 14 12:00:01 2011]

[14-Sep-2011 12:00:01] [connected at Wed Sep 14 12:00:01 2011]

[14-Sep-2011 12:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[14-Sep-2011 12:57:45] <rmatte> oh my freaking god

[14-Sep-2011 12:57:51] <rmatte> Cisco Nexus doesn't support IPSLA!?

[14-Sep-2011 12:58:06] * rmatte wants to strangle someone at Cisco

[14-Sep-2011 12:58:14] <JohnnyNOC> anyone here have ibm xservers with imms?

[14-Sep-2011 13:01:24] <Hackman238> rmatte: WTF, serisouly?

[14-Sep-2011 13:01:32] <Hackman238> JohnnyNOC: Negative.

[14-Sep-2011 13:01:39] <rmatte> yeh, apparently they'll support it sometime towards the end of the year / next year

[14-Sep-2011 13:01:50] <rmatte> this totally effs up a big IPSLA project that we're working on for a client

[14-Sep-2011 13:02:06] <rmatte> since they just installed Nexus devices at their core site, and now we can't get any IPSLA stats from them

[14-Sep-2011 13:02:09] <rmatte> unreal

[14-Sep-2011 13:02:13] <Hackman238> rmatte: Oh geez...it must be all of IOS EX

[14-Sep-2011 13:02:27] <rmatte> It's NX-OS

[14-Sep-2011 13:02:36] <rmatte> they don't even run IOS

[14-Sep-2011 13:02:38] <rmatte> lol

[14-Sep-2011 13:02:41] <Hackman238> rmatte: Wow

[14-Sep-2011 13:02:54] <rmatte> yup

[14-Sep-2011 15:10:15] <jmp242> Hi all

[14-Sep-2011 15:12:02] <psyklone> wuddup

[14-Sep-2011 15:13:03] <jmp242> not too much

[14-Sep-2011 15:20:12] <Hackman238> jmp242: Hey

[14-Sep-2011 15:20:24] <Hackman238> jmp242: How goes it?

[14-Sep-2011 15:33:00] <jmp242> Not too bad

[14-Sep-2011 15:33:11] <jmp242> how are you Hackman238

[14-Sep-2011 15:51:11] <kgoedtel> anyone else having problems with zencommand after upgrading to 3.2.0? It is running my commands if I enable debug and tail zencommand.log but the datapoints are not being populated

[14-Sep-2011 15:51:40] <kgoedtel> collector performance is showing zero datapoints for zencommand

[14-Sep-2011 15:52:00] <kocolosk> kgoedtel: yep, see the channel banner

[14-Sep-2011 15:52:14] <kocolosk> kgoedtel: one of those ptaches is for that exact issue

[14-Sep-2011 15:52:41] <kgoedtel> I am looking at the changesets in trac but not sure which from a glance

[14-Sep-2011 15:53:39] <kocolosk> kgoedtel: i believe its http://dev.zenoss.org/trac/changeset/26334

[14-Sep-2011 15:54:30] <kgoedtel> will give that a shot, thanks

[14-Sep-2011 16:36:55] <illsci> I just tried to add a location to a device and the drop down doesnt populate

[14-Sep-2011 16:37:21] <illsci> It drops down for a brief second and then never does it again and when it does it only shows one thing and it happens to fast to see what it is

[14-Sep-2011 16:37:36] <illsci> ok it says Loading...

[14-Sep-2011 16:47:26] <illsci> yet never ever displays a drop down

[14-Sep-2011 16:50:06] <rmatte> illsci: try typing / in to it

[14-Sep-2011 16:50:13] <rmatte> or /Partofthelocation

[14-Sep-2011 16:50:20] <rmatte> I believe it filters

[14-Sep-2011 16:50:35] <illsci> nope..

[14-Sep-2011 16:50:39] <illsci> it doesn't let me type anything at all

[14-Sep-2011 16:50:48] <rmatte> what browser are you using?

[14-Sep-2011 16:50:50] <kgoedtel> was this after an upgrade?

[14-Sep-2011 16:50:57] <illsci> chrome

[14-Sep-2011 16:51:03] <illsci> this is my 3.2.0

[14-Sep-2011 16:51:12] <rmatte> chrome isn't technically officially supported by Zenoss

[14-Sep-2011 16:51:15] <illsci> I've never upgraded... I just installed this version and patched the patches in the topic

[14-Sep-2011 16:51:19] <rmatte> IE/Firefox are the supported browsers

[14-Sep-2011 16:51:23] <kgoedtel> oh okay

[14-Sep-2011 16:51:23] <illsci> ok

[14-Sep-2011 16:51:25] <illsci> let me try that

[14-Sep-2011 16:51:26] <rmatte> I use firefox since it works the best

[14-Sep-2011 16:52:22] <rmatte> If you check near the bottom of the download page you'll see the level of support for each browser

[14-Sep-2011 16:52:29] <rmatte> firefox and IE are listed as full support

[14-Sep-2011 16:52:37] <rmatte> chrome, opera, and safari are listed as partial

[14-Sep-2011 16:53:04] <hmp> zenoss sux in opera btw

[14-Sep-2011 16:53:11] <rmatte> so I've heard lol

[14-Sep-2011 16:53:22] <rmatte> I use firefox because it's the most hassle free

[14-Sep-2011 16:53:29] <rmatte> pages always render the way you expect them to

[14-Sep-2011 16:53:33] <hmp> i use FF just for zenoss

[14-Sep-2011 16:53:38] <rmatte> lol

[14-Sep-2011 16:54:11] <willwh> zenoss also broken in chrome :]

[14-Sep-2011 16:54:15] <willwh> yah < FF too.

[14-Sep-2011 16:54:23] <willwh> for zenoss anyway ;]

[14-Sep-2011 16:54:34] <hmp> opera: 32 tabs of random BS ... boss walks in the room ... switch to other awesome tag ... 32 tabs of random devices

[14-Sep-2011 16:55:21] <rmatte> I use FF for everything, I can't do without greasemonkey and firebug

[14-Sep-2011 16:55:29] <rmatte> awesome plugins

[14-Sep-2011 16:56:03] <hmp> rmatte: yup, one more thing I prefer FF is web design/programming

[14-Sep-2011 16:56:07] <illsci> doesnt work under firefox either

[14-Sep-2011 16:56:08] <hmp> cant match firebug there

[14-Sep-2011 16:56:30] <rmatte> yeh

[14-Sep-2011 16:57:08] <illsci> when I click on the drop down and hold it down it shows a gray bar and a little white box over to the right

[14-Sep-2011 16:57:13] <rmatte> illsci: it could be a bug because I know that they just added that location thing in the add device dialog

[14-Sep-2011 16:57:26] <rmatte> illsci: try adding the device and specifying the location after it'd added, that should work fine

[14-Sep-2011 16:57:42] <rmatte> after it's added*

[14-Sep-2011 17:05:42] <illsci> that's actually what I'm doing

[14-Sep-2011 17:05:53] <illsci> I'm going through this Zenoss book I bought and following the examples...

[14-Sep-2011 17:06:06] <illsci> I have a bunch of devices showing up and I tried to edit that location... didn't work

[14-Sep-2011 17:08:46] <illsci> actually... I think you ahve to add in locations first

[14-Sep-2011 17:08:52] <illsci> in order to add them to a device

[14-Sep-2011 17:08:53] <illsci> maybe thats why

[14-Sep-2011 17:11:01] <illsci> yeah thats it

[14-Sep-2011 17:11:16] <illsci> YOu have to add the locations in infrastructure before you can add them to a device... makes sense...

[14-Sep-2011 17:12:14] <illsci> and now I have a dot on the dashboard map...

[14-Sep-2011 17:22:18] <rmatte> oh, you didn't even add the location

[14-Sep-2011 17:22:30] <rmatte> I assume you're quite new to Zenoss?

[14-Sep-2011 17:22:31] <rmatte> lol

[14-Sep-2011 17:23:07] <rmatte> illsci: strongly recommended reading: http://downloads.sourceforge.net/zenoss/Zenoss_Administration_06-102010-3.0-v03.pdf

[14-Sep-2011 17:28:41] kocolosk is now known as kocolosk|away

[14-Sep-2011 19:43:27] <bender183> hello

[14-Sep-2011 20:18:10] kocolosk|away is now known as kocolosk

[14-Sep-2011 20:24:51] <mrdaveh> hey all. am having an interesting one. currently i have a fresh install and have added 5 devices. i open the infrastructure pain and the main section is just sitting there saying loading

[14-Sep-2011 20:25:35] <mrdaveh> going over my logs i can see that Z2 us saying its just constantly loading the /zport/dbd/device_router page

[14-Sep-2011 20:25:45] <mrdaveh> 4 times a second

[14-Sep-2011 22:51:18] <bender183> mrdaveh, which version of zenoss ?

[14-Sep-2011 23:17:07] <mrdaveh> 3.2

[14-Sep-2011 23:26:02] <mrdaveh> and yes

[14-Sep-2011 23:26:12] <mrdaveh> running the three patches in the topic

[15-Sep-2011 00:00:01] [disconnected at Thu Sep 15 00:00:01 2011]

[15-Sep-2011 00:00:02] [connected at Thu Sep 15 00:00:02 2011]

[15-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[15-Sep-2011 01:04:30] <atelepin> Hello all! I need help with add nested relations. For example PolicyMap  to IpInterface.

[15-Sep-2011 04:24:55] <tsener> hello

[15-Sep-2011 08:11:22] klone is now known as psyklone

[15-Sep-2011 08:46:41] <jmp242> Morning

[15-Sep-2011 08:46:47] <psyklone> howdy

[15-Sep-2011 09:13:24] kocolosk is now known as kocolosk|away

[15-Sep-2011 09:19:59] <enkrypt> since upgrading to zenoss 3.2.0 (enterprise) I'm getting a lot of snmp mib errors: WARNING zen.SnmpPerfConfig: mib lookup for .1.3.6.1.4.1.2021.4.15.0 failed

[15-Sep-2011 09:27:25] <enkrypt> ah yeah sorry I see it now of the 3 zenpatches

[15-Sep-2011 09:27:37] <enkrypt> is there anywhere a  nice overview of all the zenpatches available for a certain release?

[15-Sep-2011 09:30:40] kocolosk|away is now known as kocolosk

[15-Sep-2011 09:30:51] <psyklone> enkrypt, not really a compiled list. i can tell you what the zenpatches for 3.2 are specifically for, but there's not a historical list maintained.

[15-Sep-2011 09:34:31] <enkrypt> psyklone: ok. I just migrated today, and was having problems with zen.SnmpPerfConfig not finding certain mib oid's etc, so I installed the 3 zenpatches listed in the topic. any others that I should install?

[15-Sep-2011 09:35:24] <enkrypt> psyklone: oh actually, that made it worse, now I've got other error messages in zenhub.log, about AttributeError: getOSProcessConf

[15-Sep-2011 09:36:07] <psyklone> can you dpaste a snip?

[15-Sep-2011 09:36:33] <enkrypt> coming right up

[15-Sep-2011 09:36:46] <psyklone> there's nothing in those patches that should have anything to do with that

[15-Sep-2011 09:38:02] <enkrypt> http://dpaste.com/615387/

[15-Sep-2011 09:39:45] <psyklone> hmm, two different issues there

[15-Sep-2011 09:39:58] <psyklone> did you restart zenoss after the patches?

[15-Sep-2011 09:40:12] <enkrypt> psyklone: another one here: http://dpaste.com/615389/

[15-Sep-2011 09:40:45] <psyklone> yeah, that one shows up in the first one. i think i know what that is

[15-Sep-2011 09:40:48] <psyklone> one sec

[15-Sep-2011 09:41:00] <enkrypt> psyklone: I did restart after the patches. but I am also working with distributed collectors (2 of them), which I just updated now.

[15-Sep-2011 09:41:23] <enkrypt> (and I think after an update of the remote collector he restarts the daemons, so that should be ok)

[15-Sep-2011 09:41:45] <psyklone> yeah. that second one will probably go away after the collector update

[15-Sep-2011 09:44:39] <enkrypt> the one with zSnmpVer attribute error keeps coming for now

[15-Sep-2011 09:45:00] <enkrypt> even after the hub and all collectors restarted

[15-Sep-2011 09:45:18] <enkrypt> it's related to the SNMPv3 trap handling apparently

[15-Sep-2011 09:45:23] <enkrypt> which I don't use anyway, damn

[15-Sep-2011 09:45:38] <psyklone> yeah, that one's a different ball game. give me a few minutes

[15-Sep-2011 09:47:49] <enkrypt> ok no problem take your time

[15-Sep-2011 09:57:05] <Jane_Curry> pskylone - do you have the patch number yet for the trap processing on 3.2??

[15-Sep-2011 09:57:37] <psyklone> Jane_Curry, 26337

[15-Sep-2011 09:57:46] <psyklone> topic updated to include it along with the first two

[15-Sep-2011 10:22:01] <dhopp> anybody using ext4 for RRDs?  yay or nay?

[15-Sep-2011 10:23:37] <enkrypt> dhopp: sure, don't see a reason why not. it provides some advantages over ext3

[15-Sep-2011 10:24:28] <dhopp> enkrypt that's what I thought..but I've read a few articles about not so good performance with mysql and ext4 (although a little old)

[15-Sep-2011 10:24:56] <dhopp> so I didn't know how that would apply to RRDs since the access pattern for RRDs is way different then mysql

[15-Sep-2011 10:31:11] <enkrypt> dhopp: most linux distro's are using ext4 by default now (RHEL 6 among others) so you should be ok

[15-Sep-2011 10:58:50] <nyeates> Hi all

[15-Sep-2011 10:59:20] <enkrypt> hello there

[15-Sep-2011 11:00:06] <jgartman> howdy

[15-Sep-2011 11:01:06] <themactech> Simon mentioned recently that Zenoss was going to deliver expanded ZenPack creation documentation, what is the ETA on that?

[15-Sep-2011 11:01:06] <nyeates> Welcome all to developer hour - if you have any advanced (or not advanced) questions....fire away

[15-Sep-2011 11:01:15] <dhopp> crap..I had a question that I wanted to ask and now I can't remember

[15-Sep-2011 11:01:26] * dhopp needs more sleep

[15-Sep-2011 11:01:49] <themactech> I have to deliver a ZenPack for US Department of State soon, would have been nice to have

[15-Sep-2011 11:03:35] <nyeates> themactech: We should ask Simon about this next Thur (or Fri) during the ZCA meeting. I think Simon and I will be attending during the later half and you should ask him on his plans of this. I do not know our plans in this area.

[15-Sep-2011 11:03:55] <themactech> Ok, thanks

[15-Sep-2011 11:04:23] <nyeates> I hope to be provided improved github code integration instructions in the next 2 months. Right now, our docs on this are not angled at new github users.

[15-Sep-2011 11:08:28] <nyeates> By the way, we got a new ZenPack in recently, written by a IBM systems engineer for monitoring IBM System X https://github.com/zenoss/Community-Zenpacks/tree/master/ZenPacks.community.IBMSystemxIMM

[15-Sep-2011 11:09:59] <dhopp> nyeates:  that would have been handy at my last job

[15-Sep-2011 11:09:59] <dhopp> heh

[15-Sep-2011 11:10:23] <nyeates> man, its crickets in here today.... dhopp: what you working on recently?

[15-Sep-2011 11:10:59] <dhopp> well I finally got all of my hardware together to build the new production environment out..so I have been heads down into that

[15-Sep-2011 11:11:20] <dhopp> I did a demo a couple of days for the some of the people and a question came up that I didn't know the answer to and now I can't remember it

[15-Sep-2011 11:11:22] <dhopp> doh

[15-Sep-2011 11:11:25] <enkrypt> nyeates: I'm a zenoss enterprise customer, performed my 3.1.0 to 3.2.0 upgrade today. a bit rocky at first, but psyklone helped me out. and the 3 patches listed in the topic as well.

[15-Sep-2011 11:11:39] <themactech> I need to get custom components working in my ZenPacks pronto, I have been trying for awhile with Jane's documentation but I need to land this by end of next week.  Sweating bullets right now

[15-Sep-2011 11:12:18] <dhopp> themactech:  can you post maybe a problem you are having?

[15-Sep-2011 11:12:27] <Jane_Curry> nyeates - I think that ZCA meeting is tomorrow??

[15-Sep-2011 11:12:33] <nyeates> enkrypt: great! is all up and working now? we will release a 3.2.1 maint release with fixes for those

[15-Sep-2011 11:12:43] <themactech> I need to make a ZenPack to cover Promise VTrack Raids

[15-Sep-2011 11:13:10] <nyeates> Jane_Curry: OH! Shane had told me the 23rd or so....ill clarify with him and James

[15-Sep-2011 11:13:14] <themactech> I want the Physical Drives, LUNs, Power Supplies, Controllers, and Batteries to show as components

[15-Sep-2011 11:13:28] <enkrypt> nyeates: after a manual fix instrictued by psyklone (related to the Juniper zenpack apparently) all is ok. I'm still struggling with a few VMWare ESX specific bugs, but I've got a support ticket running for that (#16052-1421)

[15-Sep-2011 11:13:30] <themactech> Can't get it to work

[15-Sep-2011 11:13:53] <themactech> Then I have to do 2 other ZenPacks where custom components are required

[15-Sep-2011 11:14:06] <dhopp> themactech:  can you elaborate on "can't get it to work"

[15-Sep-2011 11:14:07] <themactech> I have been banging my head on the wall with this for awhile

[15-Sep-2011 11:14:22] <Jane_Curry> themactech - you want to bounce some ZenPack queries off me?

[15-Sep-2011 11:14:29] <themactech> I follow Jane's documentation best as I can, my stuff does not show in components

[15-Sep-2011 11:14:33] <themactech> I would love to

[15-Sep-2011 11:14:52] <themactech> I will be working on this rest of this week and next week, minus two client visits

[15-Sep-2011 11:15:20] <themactech> Maybe some of it could be of use to you Jane, does your company only do monitoring support contracts or do you also fix the client's issues?

[15-Sep-2011 11:15:41] <themactech> One of the ZenPacks I have to make is for Raritan PDUs, great devices if you do remote support

[15-Sep-2011 11:15:52] <Jane_Curry> themactech - have you looked at some of my later ZenPacks - the ones written for ZenSystems - they all have components in them

[15-Sep-2011 11:16:25] <themactech> I was hoping to understand the process from your documentation rather than cut and paste from an existing ZenPack

[15-Sep-2011 11:16:31] <Jane_Curry> My bummer ZenPack of the week is that 3.2 breaks the MIB_Nrowser ZenPack - AGAIN!!!!

[15-Sep-2011 11:16:55] <themactech> I think your ZenPacks creation doc would have been easier if it was using something 'off the wall' instead of network ports

[15-Sep-2011 11:17:05] <pkw> enkrypt: The Juniper bug was just fixed and should be in 3.2.1

[15-Sep-2011 11:17:26] <themactech> But then again I know my python/ZenDMD/Zope kung-fu is not great

[15-Sep-2011 11:17:29] <Jane_Curry> themactech - OK understand.  I have a client meeting some of tomorrow morning but I am around the rest of the week if you want to bounce things off me

[15-Sep-2011 11:17:37] <enkrypt> pkw: ok excellent, thanks!

[15-Sep-2011 11:18:13] <Jane_Curry> pkw - is this a Juniper bug in my ZenPacks.ZenSystems.Juniper zenpack - or a different Juniper ZenPack??

[15-Sep-2011 11:18:28] <themactech> Greatly appreciated, I will try and be concise.  Only way I can pay it forward is maybe help with some diagrams for your docs, I am pretty good at making those, and I could do a French translation if you wanted

[15-Sep-2011 11:18:34] <enkrypt> ok I'm heading off, be back tomorrow, see you guys

[15-Sep-2011 11:18:34] <pkw> Jane_Curry: Enterprise Zenoss zenpack

[15-Sep-2011 11:18:51] <pkw> Jane_Curry: ZenPacks.zenoss.JuniperMonitor

[15-Sep-2011 11:18:55] <nyeates> later enkrypt

[15-Sep-2011 11:19:14] <Jane_Curry> OK - thanks - so long as it isn't me that has broken things this time

[15-Sep-2011 11:20:42] <nyeates> Devs: could something be said, very generally, about creating custom components in zenpacks? ... Im thinking, to start the convo off. I know this is an often requested and misunderstood portion of programming zenpacks

[15-Sep-2011 11:21:03] <Jane_Curry> nyeates - any chance on a comment now about the ZenPack creation documentation - especially if that meeting isn't until next week??

[15-Sep-2011 11:21:42] <jgartman> nyeates: I am ashamed, but I do not know about creating custom components.

[15-Sep-2011 11:21:57] <nyeates> Jane_Curry: no chance, because I do not know. My 1on1 with Simon is tomorrow, and from all I know, he is leading that. I'll add it to my list of questions for him

[15-Sep-2011 11:22:20] <themactech> My problem is that most of what is out there documentation-wise is more 'do step 1, then step 2, then step 3' and are usually focused on a precise example, not on the process itself

[15-Sep-2011 11:22:21] <Jane_Curry> Thanks nyeates

[15-Sep-2011 11:22:23] <jgartman> Jane_Curry: What sort of comment are you looking for? It contains best practices - how to add event details, new action types are some of the things I know are going in.

[15-Sep-2011 11:22:33] <rmatte> since this is the developer meeting I wanted to bring up some things that still need to be addressed in the new UI... it is not possible to copy multiple templates at once like it used to be in the old UI, this really slows things down when trying to copy/move large numbers of device templates between classes.  It is also impossible to move device classes (which was possible in the 2.x UI)

[15-Sep-2011 11:22:48] <rmatte> rocket said he was going to go over them with me but never ended up having time to

[15-Sep-2011 11:22:49] <themactech> I was ready Matt Ray's docs on creating a SSH ZenPack, again a very precise example

[15-Sep-2011 11:23:14] <Jane_Curry> Thanks jgartman - souns useful - my question was more focused on when we would actually see it though - guess that's Simon country

[15-Sep-2011 11:24:09] <jgartman> Jane_Curry - Keep in mind I haven't worked on it, but it's a work in progress, I don't think they have the scope nailed down completely yet.

[15-Sep-2011 11:24:14] <rocket_> jgartman: do not be ashamed for you know many other things ..

[15-Sep-2011 11:24:21] <themactech> The impact of Zenoss deployments all rest on the ability to make nice ZenPacks, as far as I am concerned, this requires custom components

[15-Sep-2011 11:24:27] <Jane_Curry> rmatte - and please can we go back to menus that say "Copy" rather than "Override"

[15-Sep-2011 11:24:47] <rmatte> rocket_: would you have time today to go over those things that we discussed the other day?

[15-Sep-2011 11:25:00] <nyeates> rmatte: seems like there is a ticket for the first of your issues....you seen that?

[15-Sep-2011 11:25:17] <rocket_> rmatte: I am trying .. I have been getting slammed ..

[15-Sep-2011 11:25:18] <rmatte> agreed, sed -i 's/Override/Copy/g' *

[15-Sep-2011 11:25:19] <pkw> themactech: Are you asking about adding custom components in general to the model or more specifically how to add those into the UI?

[15-Sep-2011 11:25:22] <themactech> I mentioned awhile back that Zenoss should offer an in-depth training session on making custom ZenPacks, this would make the community much more independent and would make the outpouring of good ZenPacks explode

[15-Sep-2011 11:25:31] <rmatte> nyeates: nah, I hadn't seen it

[15-Sep-2011 11:25:43] <Jane_Curry> The 3.2 relaese notes list "Automtic event deletion (deletes in groups of events)" - anyone know what this is about?

[15-Sep-2011 11:26:00] <rocket_> I really dont see the issue with override/copy personally .. you are overriding the template at that location .. to me that one is such a minor issue

[15-Sep-2011 11:26:04] <themactech> I have a raid chassis to monitor, it can have 16 to 96 hard drives.  I need those hard drives to show as components, with performance monitoring of my choice

[15-Sep-2011 11:26:08] <themactech> thats it

[15-Sep-2011 11:26:23] <themactech> I would then apply that to a PDU, where the components are power sockets

[15-Sep-2011 11:26:32] <rmatte> rocket_: you're not necessarily though, if the template already exists at that position you're creating a copy of it

[15-Sep-2011 11:26:55] <themactech> each showing power use, and if possible, have different level of acces, where I can designate a site admin user level that could hit a button on the UI and power cycle that device

[15-Sep-2011 11:26:59] <Jane_Curry> themactech - yeh - and I said I was prepared to run one - if there were enought takers to justify the development (or some pump-priming funding from Zenoss )

[15-Sep-2011 11:27:11] <rmatte> rocket_: and any way you look at it, you are copying the template, just because that template then takes priority over the one higher up doesn't change the terminology

[15-Sep-2011 11:27:37] <rocket_> I prefer the word override in many cases honestly

[15-Sep-2011 11:27:46] <rmatte> eugh

[15-Sep-2011 11:27:52] <rocket_> because from a device stand point I am overriding the device class below me

[15-Sep-2011 11:27:54] <Jane_Curry> rocket_ - but folk are looking for a template copy.  Especially if you are "copying" to a new destination - it's not an override

[15-Sep-2011 11:28:36] <rmatte> rocket_: it's overly confusing for people who are trying to find a way to *copy* a template and have to go hunting around only to finally figure out that that word is in use

[15-Sep-2011 11:28:38] <Jane_Curry> I have certainly had folk talk to me who were confused by the "override" terminology

[15-Sep-2011 11:28:41] <pkw> Jane_Curry: The only thing that rings a bell is that ZenDeleteHistory script was updated to allow a —unixTime option

[15-Sep-2011 11:28:45] <rocket_> that instance I can see replacing the word .. there are a couple places where there are template operations so I can understand changing some

[15-Sep-2011 11:29:08] <rocket_> rmatte: how many of those people are coming from 2.5.2?

[15-Sep-2011 11:29:20] <Jane_Curry> pkw - guess that could be it - hadn't related the two....

[15-Sep-2011 11:29:21] <rmatte> rocket_: none, I'm talking about fresh users here

[15-Sep-2011 11:29:29] <rmatte> it's a common sense issue

[15-Sep-2011 11:29:41] <rocket_> again I am listening .. I am also interjecting how I see people use it and how I explain it etc

[15-Sep-2011 11:29:46] <rmatte> you expect a simple word like "Copy" for copying a template and instead you get "Override"

[15-Sep-2011 11:30:58] <rmatte> It's like storing sugar in a container marked salt

[15-Sep-2011 11:30:59] <rmatte> lol

[15-Sep-2011 11:31:24] <iancmcc> override was chosen because it was unclear to many what aspects of a template overrode those above it. If the language is such that it's clear you're overriding any currently controlling /templates/, as opposed to copying a template, the app is not assuming that the user necessarily knows how the overriding is supposed to work in the first place.

[15-Sep-2011 11:31:25] <nyeates> Intuitively, I say the word 'copy' is much better understood by a wider group of people..... could someone make a trac ticket for this and I will try and champion it once again

[15-Sep-2011 11:31:35] <Jane_Curry> rmatte or "glucose product"

[15-Sep-2011 11:31:53] <pkw> spoke to Simon about the ZenPack creation document and he said it is still in the works but there has been a lot of work preparing this ZenPack template: https://github.com/zenoss/ZenPackTemplate

[15-Sep-2011 11:32:07] <iancmcc> it is a clear way to say "You are overriding the template" without forcing new users to go to the docs to understand the implications of /copying/ a template.

[15-Sep-2011 11:32:29] <rmatte> iancmcc: The thing is though, that's the entire basis of how Zenoss works, everything below overrides everything above, zProperties for instance, people catch on to that fact quite quickly, so using that word is just confusing because they most likely already understand that that's going to happen

[15-Sep-2011 11:32:59] * gwb235 knod

[15-Sep-2011 11:33:00] <iancmcc> It has not been our experience that most people catch on to that fact. It was quite a considered change, actually, based on feedback.

[15-Sep-2011 11:33:05] <nyeates> maybe we can use the word 'copy' for the menu command, and then explain inside the popup, what occurs

[15-Sep-2011 11:33:24] <rmatte> yeh, that's a good idea

[15-Sep-2011 11:33:51] <nyeates> iancmcc: I am interested to hear your feedback on a setup like that

[15-Sep-2011 11:34:06] <iancmcc> Verbose dialogs are a UX nightmare.

[15-Sep-2011 11:34:45] <nyeates> At the same time, I think that unclear menu items are also a UX nightmare

[15-Sep-2011 11:34:46] <iancmcc> rmatte: Power users (or just somewhat-less-new users) get it, of course, immediately, but then they aren't usually the ones who would quibble over a word choice if they understand what they want to do anyway.

[15-Sep-2011 11:34:48] <rmatte> I'm sure there's a way to make a brief one liner notification without having it be overly verbose

[15-Sep-2011 11:35:01] <rmatte> and without them having to click OK or something to get by it

[15-Sep-2011 11:35:10] <rmatte> perhaps a tooltip

[15-Sep-2011 11:35:24] <rmatte> you hover over the button and there's a very brief explanation

[15-Sep-2011 11:35:52] <iancmcc>  A tooltip explaining how templates override? That's a tough one. They are surprisingly complex. Could fall back to the docs for the whole template-name-governs bit, I suppose.

[15-Sep-2011 11:36:07] <nyeates> power users have to constantly explain and add to FAQs "how do I copy a template"

[15-Sep-2011 11:36:18] <rmatte> yeh

[15-Sep-2011 11:36:29] <rmatte> that question gets asked over and over in that exact context "Copy"

[15-Sep-2011 11:36:40] <rmatte> so it's clearly not working too well the way it is

[15-Sep-2011 11:36:40] <iancmcc> oh well, if there's a ticket it'll get looked at.

[15-Sep-2011 11:37:13] <rmatte> I mean, you could name it "The Cow Button" and once I figured out what it did I wouldn't care what it's called at that point.  I'm purely thinking about this from a new user perspective.

[15-Sep-2011 11:37:23] <nyeates> rmatte: can you make a ticket of it....ill discuss internally at a defect review and see what others think

[15-Sep-2011 11:37:30] <rmatte> yup, will do

[15-Sep-2011 11:37:43] <rmatte> though I think there have been tickets for this one already open, I'll check first

[15-Sep-2011 11:37:47] uniplexed is now known as humanix

[15-Sep-2011 11:39:26] <nyeates> we plowed that into the ground :-p stopped all other convos, sorry

[15-Sep-2011 11:39:45] <humanix> hi there

[15-Sep-2011 11:39:52] <Jane_Curry> Hi humanix

[15-Sep-2011 11:40:06] <rmatte> yeh, if anyone else has anything fire it out there

[15-Sep-2011 11:40:10] <rmatte>

[15-Sep-2011 11:40:11] <humanix> i forget my password

[15-Sep-2011 11:40:19] <rmatte> password for what?

[15-Sep-2011 11:40:19] <humanix> forgot

[15-Sep-2011 11:40:24] <humanix> freenode

[15-Sep-2011 11:40:27] <psyklone> it's **********

[15-Sep-2011 11:40:31] <humanix> lol

[15-Sep-2011 11:40:38] <rmatte> you mean for nickserv?

[15-Sep-2011 11:40:41] <humanix> yup

[15-Sep-2011 11:41:00] <Jane_Curry> Beware - zentrap in 3.2 doesn't work with SNMP v1 traps - patch 26337

[15-Sep-2011 11:41:14] <mrchippy> Mine is JALAPENOFACE

[15-Sep-2011 11:41:29] <rmatte> humanix: if you put your email address in when you registered you can probably email the freenode peeps and have them reset the password then send it to that address

[15-Sep-2011 11:41:33] <mrchippy> Sorry, inside Zenoss joke.  please ignore.

[15-Sep-2011 11:41:41] <iancmcc> mrchippy seems to be confusing this channel for internal Skype

[15-Sep-2011 11:41:59] <rmatte> Jane_Curry: wow, that zentrap one is a rough one lol

[15-Sep-2011 11:41:59] <iancmcc> he has never been the most on the ball.

[15-Sep-2011 11:42:25] <Jane_Curry> Not nice - but does now seem to be fixed

[15-Sep-2011 11:42:31] <rmatte> good stuff

[15-Sep-2011 11:42:41] <psyklone> got that out as fast as possible

[15-Sep-2011 11:42:56] <gwb235> what's the general consensus on using community zenpacks certified for 3.0 on 4.0.2.  Should they work?

[15-Sep-2011 11:43:06] <Jane_Curry> Anyone else see eroor messages in zenhub.log with thw Amazon ZenPack - see comments at end of docs/DOC-4423

[15-Sep-2011 11:43:19] <rmatte> gwb235: for 3.0 yes, not sure about 4.0.2 since Zenoss Core isn't at that version yet

[15-Sep-2011 11:43:20] <Jane_Curry> Thanks psyklone!

[15-Sep-2011 11:43:27] <iancmcc> gwb235: As long as they don't monkeypatch the events system, they will probably work.

[15-Sep-2011 11:43:50] <rmatte> gwb235: oh my bad, I see what you meant, no idea until you try lol

[15-Sep-2011 11:43:51] <gwb235> we're just doing a preview of the 4.0.2 bits - not migrating - but wanted to know what to look forward to

[15-Sep-2011 11:44:15] <gwb235> as the community zenpacks are huge in our environment (thanks all!)

[15-Sep-2011 11:44:21] <nyeates> mrchippy/JALAPENOFACE .... sweet, root access!

[15-Sep-2011 11:44:31] <rmatte> roooooooooot

[15-Sep-2011 11:44:45] <humanix> im doing it rmatte, thanks

[15-Sep-2011 11:44:48] <SDuensin> Hey guys, as a Python developer, I'm pretty dangerous.  How can I get the 'numpy' module added to my Zenoss environment?

[15-Sep-2011 11:44:53] <humanix> i got a doubt about the transformation in production states

[15-Sep-2011 11:44:57] <humanix> where can i set new states?

[15-Sep-2011 11:45:19] <iancmcc> SDuensin: $ZENHOME/bin/easy_install numpy

[15-Sep-2011 11:45:27] <rmatte> humanix: you mean you want to set a new state in the transform code?

[15-Sep-2011 11:45:27] <SDuensin> iancmcc, thanks!

[15-Sep-2011 11:45:39] <humanix> yup

[15-Sep-2011 11:45:49] <humanix> exactly rmatte

[15-Sep-2011 11:46:15] <rmatte> humanix: evt.prodState = 1000

[15-Sep-2011 11:46:23] <rmatte> where 1000 is the value of the state...

[15-Sep-2011 11:46:49] <rmatte> you can get the values in Advanced

[15-Sep-2011 11:47:01] <rmatte> under State Conversions field

[15-Sep-2011 11:49:45] <humanix> rmatte i made it

[15-Sep-2011 11:49:46] <humanix> thanks

[15-Sep-2011 11:49:47] <humanix>

[15-Sep-2011 11:49:51] <rmatte> np

[15-Sep-2011 11:50:58] <humanix> done

[15-Sep-2011 11:51:02] <humanix> added to automatic

[15-Sep-2011 11:51:07] <Jane_Curry> themactech - go and look at the  ZenPack template referenced earlier (if you haven't already done so) : https://github.com/zenoss/ZenPackTemplate

[15-Sep-2011 11:51:20] <Jane_Curry> Looks like very helpful stuff

[15-Sep-2011 11:51:55] <Jane_Curry> OK - can I open questions on reports again???

[15-Sep-2011 11:52:14] <themactech> Just got off phone with client, now I have to read 10 pages of chat...

[15-Sep-2011 11:52:48] <jgartman> If you reference the zenpack template please please understand that it's not complete.

[15-Sep-2011 11:53:05] <nyeates> jane_curry: That EC2Monitor stuff is curious....you said that you are using default device classes (/Server/Linux and /Server/Windows). Do you mean seperately, the ec2 instances are also being monitored here by like snmp instead of via cloudwatch apis, which is what ec2monitor zp uses

[15-Sep-2011 11:56:03] <Jane_Curry> nyeates - the real devices discoverd by the AWS ZenPack into /Server/Linux and /Server/Windows are monitored by SNMP but  modeleing my EC2Manager device has

[15-Sep-2011 11:56:24] <Jane_Curry> a message saying SNMP minitoring off for EC2Manager

[15-Sep-2011 11:57:22] <Jane_Curry> .. although ConfigureProperties for EC2Manager has zSnmpMonitorIgnore = False

[15-Sep-2011 11:58:07] <Jane_Curry> jgartman - understand the zenpack templatre isn't complete - still some good hints in there

[15-Sep-2011 11:58:26] <rmatte> Voila: http://dev.zenoss.org/trac/ticket/7867

[15-Sep-2011 11:59:06] <Jane_Curry> In 3.2 with patch 26337 to fix TRAPS - why has severity default changed from Warning to Info - or is this another bug??

[15-Sep-2011 11:59:56] <nyeates> I can help you do some troubleshooting by calling zencw2 manually

[15-Sep-2011 12:00:01] [disconnected at Thu Sep 15 12:00:01 2011]

[15-Sep-2011 12:00:02] [connected at Thu Sep 15 12:00:02 2011]

[15-Sep-2011 12:00:15] [zenoss-logger (logger bot) has joined #zenoss]

[15-Sep-2011 12:00:39] <themactech> I assume all the 3.2 patches will be rolled into a 3.2.1 at some point, I plan to stick to 3.1 until then, is that a decent plan?

[15-Sep-2011 12:01:22] <Jane_Curry> themactech - I would say so - don't go there yet

[15-Sep-2011 12:01:39] <psyklone> Jane_Curry, that patch doesn't affect severity

[15-Sep-2011 12:01:42] <rmatte> iancmcc: in terms of a brief tooltip, I came up with a potential one and added it to the ticket: "Copied templates will override identically named templates higher up."

[15-Sep-2011 12:01:50] <rmatte> I think that pretty much says it all in one sentence

[15-Sep-2011 12:01:52] <Jane_Curry> Consistent event consoles ie. you have filter boxes in all event consoles - is jolly useful

[15-Sep-2011 12:01:54] <iancmcc> i got it

[15-Sep-2011 12:02:01] <rmatte> k

[15-Sep-2011 12:02:04] <nyeates> themactech: yes, we will be doing a 3.2.1 beta very soon.... let it run for a week or two, and then release 3.2.1 which should cover many of these bugs that have been seen

[15-Sep-2011 12:02:35] <Jane_Curry> psyklone - in that cae, something else has changed sev down to info

[15-Sep-2011 12:04:05] <nyeates> Dev Chat officially over everyone...thanks for participating!

[15-Sep-2011 12:04:15] <rmatte> thanks for your time guys

[15-Sep-2011 12:04:26] <Jane_Curry> thanks nick

[15-Sep-2011 12:06:24] <nyeates> Jane_Curry: let me work with you in those comments (so its public) and here to fix ec2....i have some experience there...youll have to give me some time now to settle though

[15-Sep-2011 12:07:24] <dhopp> is it possible to disable the "network map" feature?

[15-Sep-2011 12:08:02] <Jane_Curry> nyeates - no worries -  here for a couple of hours yet

[15-Sep-2011 12:08:47] <rmatte> dhopp: no, you can't outright disable it

[15-Sep-2011 12:08:58] <rmatte> dhopp: though it is probably the most useless feature at the moment

[15-Sep-2011 12:09:07] <rmatte> I can't remember the last time I bothered looking at it

[15-Sep-2011 12:12:59] <dhopp> rmatte: I'm not worried about me..I'm worried about people that like eye candy

[15-Sep-2011 12:13:06] <dhopp> and at initial glance it does look pretty cool

[15-Sep-2011 12:13:47] <rmatte> yeh, until you start using it lol

[15-Sep-2011 12:15:29] <dhopp> rmatte:  I can just see people that don't really understand networks trying to use during a problem and going "look this other network is connected over here..could that be causing a problem" and then me or whoever is working on problem having to waste time denying it

[15-Sep-2011 12:16:32] <Jane_Curry> I find the little tracepath utility is more useful than the network map often

[15-Sep-2011 12:21:13] <rmatte> there's a tracepath utility?

[15-Sep-2011 12:21:48] <rmatte> oh, you mean the actual command?

[15-Sep-2011 12:22:03] <rmatte> (thought you meant it was a built in feature in Zenoss)

[15-Sep-2011 12:27:16] <nyeates> Jane_Curry: I have to leave for a bit. I will be back on later to assist with ec2....look around for zencw2 for now....i couldnt find it when i looked in $ZENHOME/bin

[15-Sep-2011 12:28:02] <bigegor> hi all

[15-Sep-2011 12:28:03] <nyeates> basically zencommand runs zencw2 which includes libraries that call on Cloudwatch calls (amazons monitoring API)

[15-Sep-2011 12:29:12] <dhopp> hi bigegor

[15-Sep-2011 12:31:32] <bigegor> it looks like Zenoss 3.2 don't refreshed  RRD Graphs, manual refresh works.

[15-Sep-2011 12:34:19] <rmatte> bigegor: ah, so no auto-refresh, fun

[15-Sep-2011 12:35:00] <Jane_Curry> no - tracepath is something I found on the forums way back - wouter I think

[15-Sep-2011 12:35:07] <rmatte> ah

[15-Sep-2011 12:35:33] <Jane_Curry> nyeates - do I need toactively enable clodwatch on thye Amazon side?

[15-Sep-2011 12:35:52] <bigegor> it possible now to set refresh interwal, but it doesn't work as expected

[15-Sep-2011 12:36:31] <rmatte> bigegor: I'll test it to confirm then we can open a trac ticket (if you haven't already)

[15-Sep-2011 12:37:05] <rmatte> hmmm, my 3.2 install is misbehaving, it just sits at "Loading. Please wait..." when I click on Infrastructure

[15-Sep-2011 12:37:40] <rmatte> hmmm, actually might be a browser cache issue

[15-Sep-2011 12:41:48] <rmatte> nope, cleared cache and still doing it, will try restarting Zenoss

[15-Sep-2011 12:42:22] <Jane_Curry> bigegor - you serious?  No graphs refresh automatically?

[15-Sep-2011 12:43:11] <rmatte> that's what he's saying yeh, I'm trying to get my 3.2 to load my infrastructure page so that I can confirm

[15-Sep-2011 12:43:26] <rmatte> I just applied the 3 patches for the first time

[15-Sep-2011 12:43:41] <Jane_Curry> rmatte - I saw some of that coupla days ago - have a look in event.log - I had a message about an http message being too big - Hackman helped me resolve

[15-Sep-2011 12:43:55] <rmatte> ah ok, I'll take a look

[15-Sep-2011 12:44:03] <bigegor> another funny things, is datapoints conter will be reseted after 3-4 collection cycles (tested with zenperfsql daemon)

[15-Sep-2011 12:44:30] <rmatte> you talking about counter type datasources?

[15-Sep-2011 12:45:34] <bigegor> datapoints counter in log

[15-Sep-2011 12:45:46] <rmatte> Jane_Curry: I still have the same issue but no messages in event.log at all

[15-Sep-2011 12:45:49] <nyeates> Jane_Curry: it is my understanding that our EC2Monitor ZP turns CloudWatch on if its off. Also, I think that Cloudwatch became free (that needs verification). It might be a good idea to turn it on manually on the AWS side, just to reduce the variables

[15-Sep-2011 12:46:14] <rmatte> nyeates: so I've run in to a pretty serious issue with my 3.2 test box, but I have no idea how i'd go about determining the source

[15-Sep-2011 12:46:15] <Jane_Curry> This was the message I had

[15-Sep-2011 12:46:20] <Jane_Curry> 2011-09-13T19:43:04 ERROR ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 127.0.0.1:57215 at 0xab14b4c channel#: 75 requests:> (<type 'exceptions.ValueError'>:HTTP headers invalid (too long) (got: 8227 bytes, allowed 8192 bytes [/usr/local/zenoss/python/lib/python2.6/asyncore.py|read|74] [/usr/local/zenoss/python/lib/python2.6/asyncore.py|handle_read_event|413] [/usr/local/zenoss/python/lib/

[15-Sep-2011 12:46:22] <Jane_Curry> python2.6/asynchat.py|handle_read|155] [/usr/local/zenoss/python/lib/python2.6/site-packages/Zope2-2.12.1-py2.6-linux-i686.egg/ZServer/HTTPServer.py|collect_incoming_data|435])

[15-Sep-2011 12:46:32] <rmatte> when I click on Infrastructure, it goes in to "Loading. Please wait..." dialog forever

[15-Sep-2011 12:46:47] <rmatte> I'll try from a different computer

[15-Sep-2011 12:47:45] <Jane_Curry> My solution was to hack zope.conf and uncomment the line with http-header-max-length 16384 and recycle Zope

[15-Sep-2011 12:47:49] <rmatte> hmmm, strange it actually works on the other system, ok there must be something else to do with my browser

[15-Sep-2011 12:48:09] <rmatte> nah, don't think it's a server issue in this particular case

[15-Sep-2011 12:48:40] <rmatte> anyways, I'll just use the other system for now and fix this later

[15-Sep-2011 12:48:41] <Jane_Curry> Clear all cache AND all cookies - I had a hosed process display and clearing cookies solved that

[15-Sep-2011 12:49:05] <rmatte> I hate clearing my cookies because it clears my banking cookie then I have to re-enter a bunch of stuff lol

[15-Sep-2011 12:49:08] <rmatte> but guess I'll have to

[15-Sep-2011 12:49:47] <rmatte> ok, cleared everything...

[15-Sep-2011 12:49:53] <nyeates> sometimes you can manually clear cookies

[15-Sep-2011 12:50:23] <rmatte> nope, still doing it

[15-Sep-2011 12:50:29] <rmatte> and I've cleared absolutely everything

[15-Sep-2011 12:50:38] <rmatte> no messages in event.log

[15-Sep-2011 12:51:03] <rmatte> it only does it on that one page

[15-Sep-2011 12:51:38] <Jane_Curry> nyeates - I seem to have detailed monitoring ON on all my instances and I certainly didn't turn it on manually so it must be the ZP done that

[15-Sep-2011 12:51:57] <nyeates> rmatte: kill all, start daemons one at a time and watch log or screen output ....make sure to start with -v10 or verbose

[15-Sep-2011 12:52:18] <nyeates> sucks to need to do that, but when there are no error messages, you sometimes gotta

[15-Sep-2011 12:52:21] <rmatte> It's annoying, it actually loads the devices on the screen but then I get an overlay overtop of them with the loading dialog so I can't actually click on any of them

[15-Sep-2011 12:52:40] <rmatte> nyeates: it actually works from a different computer though

[15-Sep-2011 12:52:53] <rmatte> I have a windows VM running on top of Linux and it works on that

[15-Sep-2011 12:52:58] <rmatte> strange

[15-Sep-2011 12:53:03] <nyeates> weird

[15-Sep-2011 12:53:45] <rmatte> anyways, I'll test the graph reload on that for now

[15-Sep-2011 12:53:58] <nyeates> Jane_Curry: did u happen to find zencw2 ?

[15-Sep-2011 12:54:03] <nyeates> if not, ill go hunting for it

[15-Sep-2011 12:54:03] <rmatte> I should open up firebug and see what's going on with that loading thing

[15-Sep-2011 12:54:36] <rmatte> nyeates: at some point you guys need to add the ability to specify an exact range for graphs

[15-Sep-2011 12:54:39] <rmatte> it's a gaping feature

[15-Sep-2011 12:54:46] <rmatte> or lack thereof

[15-Sep-2011 12:54:58] <rmatte> as in "this date to this date"

[15-Sep-2011 12:55:31] <rmatte> I know that rackspace have their own ZenPack that adds it, but it's kind of ridiculous to have a third party pack just for that

[15-Sep-2011 12:55:39] <nyeates> Jane_curry: ok here /opt/zenoss/ZenPacks/ZenPacks.zenoss.ZenAWS-1.0.3-py2.6.egg/ZenPacks/zenoss/ZenAWS/libexec/zencw2.py

[15-Sep-2011 12:55:58] <nyeates> let me figure out how to run it again, i forget

[15-Sep-2011 12:56:20] <rmatte> hmmm, I just noticed something

[15-Sep-2011 12:56:36] <rmatte> the graph reload value was defaulted to 30 minutes

[15-Sep-2011 12:56:54] <bigegor> have somebody experience with debuging python + external C libraries (pysamba)?

[15-Sep-2011 12:56:55] <rmatte> actually, it's not defaulted to anything

[15-Sep-2011 13:01:36] <rmatte> well, egor's right about the graphs not refreshing after 5 minutes by default which they should...

[15-Sep-2011 13:01:45] <rmatte> I'm seeing if I actually select 5 minutes if it'll do it

[15-Sep-2011 13:01:55] <rmatte> could be that it's just not assigning a default refresh value

[15-Sep-2011 13:01:59] <nyeates> Jane_Curry: Run this python zencw2.py --help

[15-Sep-2011 13:02:59] <rmatte> also, the refresh values in the dropdown aren't the greatest, instead of having 30 minutes and an hour as options it'd be nice to have a 1 minute option in there

[15-Sep-2011 13:03:56] <nyeates> Jane_Curry: and make a command something like: python zencw2.py -u 1234567890 -p 882384KA0092349LALflsd EC2Instance

[15-Sep-2011 13:04:07] <Hackman238> Hello all

[15-Sep-2011 13:04:14] <Hackman238> nyeates: Can you op me?

[15-Sep-2011 13:04:16] <nyeates> bu with your real AWS userkey and privatekey

[15-Sep-2011 13:04:17] <rmatte> Hackman238: aloha mon ami

[15-Sep-2011 13:04:39] <Hackman238> rmatte: Hey. Sorry I missed the dev meeting. Day errands to do today

[15-Sep-2011 13:04:40] <rmatte> I think we lost our ops during that server restart yesterday

[15-Sep-2011 13:04:46] <rmatte> Hackman238: all good

[15-Sep-2011 13:05:02] <rmatte> we're finding new bugs in 3.2

[15-Sep-2011 13:05:09] <nyeates> I thought i gave u guys auto ops

[15-Sep-2011 13:05:13] <rmatte> bigegor just brought it to our attention that graphs don't auto-refresh

[15-Sep-2011 13:05:13] <bigegor> rmatte: not it doesn't. I've tried 5 min too.

[15-Sep-2011 13:05:15] <rmatte> and he's correct

[15-Sep-2011 13:05:46] <rmatte> nyeates: ah, I wasn't identified with nickserv, that might have been why

[15-Sep-2011 13:05:53] <nyeates> great - thx for finding this stuff guys

[15-Sep-2011 13:05:58] <nyeates> rmatte: yeh u need that

[15-Sep-2011 13:06:38] <rmatte> bigegor: did you want to open the trac ticket for this one or shall I?

[15-Sep-2011 13:07:17] <nyeates> announce that trac url here when its made please, i want to follow it

[15-Sep-2011 13:07:21] <rmatte> I haven't opened trac tickets in ages since I can't really open tickets when I'm primarily using an old version

[15-Sep-2011 13:07:21] <rmatte> lol

[15-Sep-2011 13:07:33] <rmatte> back when we were using cutting edge I was opening them all the time

[15-Sep-2011 13:07:49] <rmatte> nyeates: will do

[15-Sep-2011 13:08:00] <bigegor> rmatte: please, open ticket.

[15-Sep-2011 13:08:14] <rmatte> ok, doing it now...

[15-Sep-2011 13:09:37] <nyeates> Hackman238: when is ZCA meeting? jane said tomorrow

[15-Sep-2011 13:09:43] <rmatte> it is tomorrow, correct

[15-Sep-2011 13:10:17] <nyeates> Hackman238: what happened to the 9/23 meeting we discussed about me and Simon hoping in on? do i have my dates wrong?

[15-Sep-2011 13:11:34] <nyeates> Jane_Curry: when i put in horribly fake userkey and privatekey data, i get 404 errors....not sure if this is cause of the made up keys, or if something is wrong

[15-Sep-2011 13:12:41] <kgoedtel> is there any way to copy datasources or graphs from one template to another?

[15-Sep-2011 13:12:52] <kgoedtel> or even clone them within a template

[15-Sep-2011 13:14:09] <nyeates> short of programming something in the DMD to do this....oh wait, maybe yeah

[15-Sep-2011 13:14:10] <rmatte> nyeates: http://dev.zenoss.org/trac/ticket/7868

[15-Sep-2011 13:14:47] <nyeates> you *might* be able to use the ZMI, Zope Management Interface, to copy stuff. You simply add a "/manage" to any url in zenoss

[15-Sep-2011 13:15:39] <kgoedtel> would make my life sooo much easier than re-entering 90% of the same thing for a hundred different datasources

[15-Sep-2011 13:16:01] <Hackman238> nyeates: We're going to have another

[15-Sep-2011 13:16:12] <Hackman238> This meeting is off cycle

[15-Sep-2011 13:16:15] <Hackman238> The next isnt

[15-Sep-2011 13:17:29] <nyeates> Hackman: does that mean no one will be at the next one?

[15-Sep-2011 13:17:37] <nyeates> or rather, less people

[15-Sep-2011 13:17:38] <Hackman238> nexex: negative

[15-Sep-2011 13:17:42] <Hackman238> nyeates: negative

[15-Sep-2011 13:18:26] <nyeates> im just thinking....people attend one week, very next week another meeting "eh, i wont miss much"

[15-Sep-2011 13:19:12] <nyeates> We will just be sure to advertise that next week Simon and I are coming and giving updates on X,Y,Z

[15-Sep-2011 13:20:12] <Hackman238> nyeates: Not really, the folksa rent really like that

[15-Sep-2011 13:20:29] <Hackman238> nyeates: It'll also help control the meeting length since we have much to catch up on

[15-Sep-2011 13:20:40] <nyeates> kgoedtel: try http://<yourzenoss>:8080/zport/dmd/manage

[15-Sep-2011 13:21:03] <nyeates> kgoedtel: be VERY careful in here though, as you can fubar stuff easily

[15-Sep-2011 13:21:35] <nyeates> try to find the area where datasource is stored in the hierarchy, and then there are often check boxes next to objects and there are copy and paste buttons

[15-Sep-2011 13:21:39] <Jane_Curry> nyeates tried your command with EC2Instance on the end and just get a prompt back - should that be a real instance id???

[15-Sep-2011 13:22:12] <nyeates> Hackman238: true on the meeting length....so which day and time next week for sure? is it known?

[15-Sep-2011 13:22:37] <Hackman238> nyeates: Let me check with jmp, he's the scheduling guy

[15-Sep-2011 13:22:50] <Hackman238> nyeates: God....I'm so screwed on time right now.

[15-Sep-2011 13:22:52] <Hackman238> LOL

[15-Sep-2011 13:23:15] <nyeates> u didnt know, its midnight!

[15-Sep-2011 13:23:26] <Jane_Curry> nyeates: I also get a screed of error messages with a bad key

[15-Sep-2011 13:23:54] <Jane_Curry> but null response with a good key

[15-Sep-2011 13:24:05] <Hackman238> nyeates: So many projects...After reviewing the restore model of Windows 8 I'm going to ahve to rethink my retail business models

[15-Sep-2011 13:24:23] <nyeates> Jane_Curry: feel free to try around with other things in the command.....i had said EC2Instance, as if you just enter it with nothing, it says:

[15-Sep-2011 13:24:23] <nyeates> zencw2.py command [options] need to have command EC2Manager, EC2Instance or EC2InstanceType

[15-Sep-2011 13:24:50] <Hackman238> nyeates: Additionally planned to bring my R  &  D office in downtown San Antonio online the 30th, and just recieved a few large RFP's.

[15-Sep-2011 13:25:11] <nyeates> win 8 restore model? do enlighten us

[15-Sep-2011 13:25:50] <Hackman238> nyeates: Oh man....its so simple to parallel install a mongoliod could do it

[15-Sep-2011 13:25:57] <rmatte> apparently the UI in Windows Server 8 is optional, you'll be able to manage it completely via commandline

[15-Sep-2011 13:26:06] <rmatte> nice to see that microsoft are finally learning

[15-Sep-2011 13:26:15] <Hackman238> nyeates: I coulnt break Windows 8 in a way that the built in auto repair systems coulnt fix it in less than 10 minutes

[15-Sep-2011 13:26:25] <Hackman238> nyeates: Even intentionally

[15-Sep-2011 13:26:50] <rmatte> lol

[15-Sep-2011 13:26:54] <rmatte> format c:

[15-Sep-2011 13:26:56] <rmatte> broken

[15-Sep-2011 13:27:06] <Jane_Curry> I think the problem is in the zenhub.log message - it says it can't import 'ImportError: No module named zenoss.ZenAWS.parsers.ec2.instances\')\n'

[15-Sep-2011 13:27:27] <Jane_Curry> the thing is in ZenPacks.zenoss.ZenAWS.parsers.ec2.instances

[15-Sep-2011 13:27:44] <Jane_Curry> If you use zendmd to import zenoss.ZenAWS.parsers.ec2.instances then it fails

[15-Sep-2011 13:27:53] <jmp242> Hackman238: never challenge someone

[15-Sep-2011 13:27:59] <jmp242> on breaking an MS product lol

[15-Sep-2011 13:28:00] <Hackman238> rmatte: I deleted everything I could in windows...rebooted...5 mintues later it was at the damn logon screen

[15-Sep-2011 13:28:04] <Jane_Curry> If you use zendmd to import ZenPacks.zenoss.ZenAWS.parsers.ec2.instances then it works

[15-Sep-2011 13:28:12] <jmp242> Hackman238: run DBAN lol

[15-Sep-2011 13:29:27] <Hackman238> jmp242: LOL

[15-Sep-2011 13:29:37] <Hackman238> Anyway, need to rethink our restoration model

[15-Sep-2011 13:30:54] <Jane_Curry> nyeates: Null response with any of those 3 parameters to zencw2.py

[15-Sep-2011 13:31:35] <nyeates> Jane_Curry: interesting. let me go get aws keys and give it a try on a 3.2 system.... ill get back to u

[15-Sep-2011 13:32:06] <nyeates> Jane_Curry: are you think that an import in the code somewhere is maybe missing the ZenPacks. ?

[15-Sep-2011 13:38:18] <Jane_Curry> nyeates: The error message is suggesting that and my zendmd tests support it

[15-Sep-2011 13:45:02] <jmp242> Hmmm, I don't even want to think about Windows 8 lol

[15-Sep-2011 13:45:52] <Hackman238> jmp242: Well half my income comes from my two retail stores

[15-Sep-2011 13:46:15] <Hackman238> jmp242: We sell 'restoration' contracts for PCs

[15-Sep-2011 13:46:53] <rmatte> what do "restoration contracts" entail?

[15-Sep-2011 13:47:26] <Hackman238> jmp242: Results in highly predictable and reliable income, highly efficient detritus free restorations, nearly 95% 'exceptionally' satisfied customers who advertise for us

[15-Sep-2011 13:49:16] <Hackman238> rmatte: You bring in your PC. Its in need of a restoration. Two options- 1) Restore it, no contract. Its expensive and offers no warrenty, etc. 2) Smart Restoration contract - We restore the device, for the restoration fee, and you buy the contract. The contract period allows you to bring that same device in for unlimited re-restorations using our saved initial restoration image of your device for the contract period, each restore only cos

[15-Sep-2011 13:50:28] <Hackman238> rmatte: We do not remove viruses, except in exceptional situations. you want it fixed, we'll restore it. You want it to remaind working (especially if its your kids machines) buy the contract and drop if off as often as you like. Often same business day service for contract restorations

[15-Sep-2011 13:51:33] <rmatte> ohhh

[15-Sep-2011 13:51:41] <rmatte> so you're essentially a ghost shop

[15-Sep-2011 13:52:01] <rmatte> your first sentence got cut off there...

[15-Sep-2011 13:52:04] <rmatte> "each restore only cos"

[15-Sep-2011 13:52:10] <Hackman238> rmatte: This does a few things - 1) my staff are specialized in hardware troubleshooting and our restoration software. Low cost staff 2) Staff are not idle waiting for scans, etc 3) We can target and please the low income and volume crowds 4) we do not advertise because customers machines are more often working than not, for what appears to be less to them. They advertise to their friends

[15-Sep-2011 13:52:39] <Hackman238> rmatte: ...each restore only costing  a very low fee.

[15-Sep-2011 13:52:45] <rmatte> ah

[15-Sep-2011 13:53:19] <rmatte> that's actually a pretty damn good idea

[15-Sep-2011 13:53:21] <Hackman238> rmatte: The idea is this- minimize cost, maximize service, capitalize on a market that is unconquered

[15-Sep-2011 13:53:32] <Hackman238> rmatte: I make a god damn fortune

[15-Sep-2011 13:53:36] <Hackman238>

[15-Sep-2011 13:53:42] <rmatte> I know there's nothing like that around here, I think I may start something up

[15-Sep-2011 13:53:47] <rmatte> I've been looking for a business venture

[15-Sep-2011 13:54:05] <rmatte> do you actually run this out of a store?

[15-Sep-2011 13:54:08] <rmatte> or just someone'

[15-Sep-2011 13:54:13] <rmatte> erm, someone's basement

[15-Sep-2011 13:54:25] <Hackman238> rmatte: If a customer wants virus removal we tell them to go to frys (for the CA office) or BestBuy (for the NY office). We tell them they'll eventually come back and spend less with us once they are tired of Fing around all the time

[15-Sep-2011 13:54:35] <Hackman238> I have two retail stores

[15-Sep-2011 13:54:44] <rmatte> nice

[15-Sep-2011 13:54:51] <Hackman238> San Marcos, CA and Watervliet, NY

[15-Sep-2011 13:55:05] <Hackman238> Plus my consulting business, day job and R  &  D ops

[15-Sep-2011 13:55:08] <Hackman238> LOL

[15-Sep-2011 13:55:17] <rmatte> nice, good setup

[15-Sep-2011 13:55:29] <rmatte> once of my friends does consulting and he's going to start throwing some work my way

[15-Sep-2011 13:55:36] <rmatte> pays $40/hour about, which isn't bad

[15-Sep-2011 13:56:05] <rmatte> I did 3 hours of work for him over the weekend since he was gone to a funeral

[15-Sep-2011 13:56:06] <Hackman238> rmatte: If you're interested in my software that makes my stores run, let me know. If I thinnk your area can support a business and I think you can manage it, I'll invest

[15-Sep-2011 13:56:20] <Hackman238> rmatte: Nice

[15-Sep-2011 13:56:36] <rmatte> well, I have some idea of how I'd set it up

[15-Sep-2011 13:56:42] <rmatte> I mean, there's lots of imaging software out there

[15-Sep-2011 13:56:47] <rmatte> just need some major storage

[15-Sep-2011 13:56:56] <Hackman238> rmatte: Negative- dont do that. You'll need staff that think too much

[15-Sep-2011 13:57:06] <rmatte> and then I'd assign each customer an ID number stored in a database, and store the images under those numbers

[15-Sep-2011 13:57:15] <rmatte> and potentially encrypt the data for security reasons

[15-Sep-2011 13:57:32] <Hackman238> rmatte: My AutoAll is all dummy proof and menu driven. It even backs up and restores software and directories automagically

[15-Sep-2011 13:57:41] <rmatte> nah, I could write up some integration code to make the process braindead if I wanted to

[15-Sep-2011 13:57:44] <rmatte> but...

[15-Sep-2011 13:57:58] <rmatte> if you are interested in something like that here we can discuss

[15-Sep-2011 13:58:05] <Hackman238> rmatte: If they staff think too much, they'll try to 'fix' machines that should just be images

[15-Sep-2011 13:58:07] <rmatte> this is a very IT oriented town, lots of computer business

[15-Sep-2011 13:58:17] <rmatte> but I've yet to see a service like that offered residentially

[15-Sep-2011 13:58:18] <Hackman238> rmatte: Anything like my model there?

[15-Sep-2011 13:58:27] <Hackman238> rmatte: Gotcha

[15-Sep-2011 13:58:41] <rmatte> The big thing would be advertising to get the business in

[15-Sep-2011 13:58:53] <rmatte> what's your business actually called?

[15-Sep-2011 13:58:54] <Hackman238> rmatte: Nein nein, dont need to

[15-Sep-2011 13:59:13] <rmatte> well, how do people find out about it if you don't do a bit of advertising at the start?

[15-Sep-2011 13:59:32] <Hackman238> rmatte: Use free advertising sources and offer initial 'grand opening' contracts that are very inexpensive

[15-Sep-2011 13:59:48] <Hackman238> rmatte: Low ball to under cut everyone.

[15-Sep-2011 14:00:11] <Hackman238> rmatte: Once you have a few hundred customers spreading the word, end the intro offer

[15-Sep-2011 14:00:22] <kgoedtel> give them a free razor so they buy more blades

[15-Sep-2011 14:00:30] <Hackman238> kgoedtel: Thats right

[15-Sep-2011 14:00:42] <Hackman238> kgoedtel: Thats what Gillet did and it works

[15-Sep-2011 14:00:54] <rmatte> I'm not talking about million dollar ads, for instance, we have a free computer magazine called "The Monitor" which is available in most computer stores around here for free.  It's been around for ages, I used to get BBS numbers out of the back of it.  Just a cheap little ad in that would go a long way.

[15-Sep-2011 14:01:14] <Hackman238> rmatte: Oh yeah, those are great places if the local area has them

[15-Sep-2011 14:01:34] <rmatte> exactly

[15-Sep-2011 14:01:46] <Hackman238> rmatte: The idea is to keep the op lean cost, efficient and aim for 100% satisfaction

[15-Sep-2011 14:01:56] <rmatte> yup

[15-Sep-2011 14:02:08] <Hackman238> rmatte: A  &  G and J  &  B

[15-Sep-2011 14:02:28] <rmatte> ?

[15-Sep-2011 14:02:31] <rmatte> that's the name of it/

[15-Sep-2011 14:02:31] <Hackman238> rmatte: Both were started independantly, I bought in and brought a new bmodel

[15-Sep-2011 14:02:32] <rmatte> ?

[15-Sep-2011 14:02:36] <Hackman238> Yep

[15-Sep-2011 14:02:37] <rmatte> ah

[15-Sep-2011 14:02:53] <Hackman238> Jim in CA, John in NY

[15-Sep-2011 14:03:02] <rmatte> if you'd open another would you be aiming to stick with that naming?

[15-Sep-2011 14:03:03] <rmatte> lol

[15-Sep-2011 14:03:14] <Hackman238> rmatte: Call it whatever you want

[15-Sep-2011 14:03:20] <Hackman238> rmatte: Name doesnt matter

[15-Sep-2011 14:03:25] <rmatte> k lol

[15-Sep-2011 14:03:34] <Hackman238> rmatte: All ameri-fools remember is it was cheap and their shit works

[15-Sep-2011 14:03:42] <rmatte> yeh

[15-Sep-2011 14:03:55] <Hackman238> rmatte: And frankly they do end up paying less in many cases

[15-Sep-2011 14:03:59] <rmatte> well, this would obviously be a Canadian branch

[15-Sep-2011 14:04:09] <Hackman238> rmatte: Canadi-fool

[15-Sep-2011 14:04:13] <rmatte> hehe

[15-Sep-2011 14:04:16] <Hackman238> rmatte: I imagine we're not much differnt

[15-Sep-2011 14:04:22] <rmatte> probably not

[15-Sep-2011 14:04:22] <Hackman238> LOL

[15-Sep-2011 14:04:57] <Hackman238> Anyway, this is all way of topic for the Zenoss thread. Email me if you're really serious and we'll put a call together to chat

[15-Sep-2011 14:05:06] <rmatte> k cool

[15-Sep-2011 14:05:11] <Hackman238> Anyone else interested, contact me as well. Hackman238@gmail.com

[15-Sep-2011 14:05:26] <Hackman238> I gotta get back to work. LOL

[15-Sep-2011 14:05:29] <rmatte> same

[15-Sep-2011 14:05:47] <Hackman238> Anywho, the long and short of it is Windows 8 requires a model change-

[15-Sep-2011 15:00:56] <Hackman238> Quiet in here

[15-Sep-2011 15:01:01] <rmatte> fairly

[15-Sep-2011 15:03:38] <jmp242> netbackup fun

[15-Sep-2011 15:03:46] <jmp242> getting the new version and navigating Symantec's website

[15-Sep-2011 15:03:53] <jmp242> can be a full time job

[15-Sep-2011 15:04:56] <SDuensin> Hackman238, been writin' code!

[15-Sep-2011 15:39:49] <nyeates> sweet - 'Override template' will finally say 'Copy / Override template'

[15-Sep-2011 15:41:10] <dhopp> nyeates:  YAY!

[15-Sep-2011 15:41:21] <dhopp> is that in 3.2.1 ;-)

[15-Sep-2011 15:59:58] <humanix> hi there

[15-Sep-2011 16:00:09] <humanix> i have a doubt about monitoring a 10.000gbps switch

[15-Sep-2011 16:00:16] <humanix> with zenoss 2.5.2

[15-Sep-2011 16:00:27] <humanix> the number of packages of the default template seems to be ok

[15-Sep-2011 16:00:42] <humanix> but the throughtput graph shows me like the 10% of consumption

[15-Sep-2011 16:00:49] <humanix> what can this be?

[15-Sep-2011 16:00:56] <nyeates> dhopp: it will come in next major release

[15-Sep-2011 16:01:15] <rmatte> humanix: make sure you're monitoring the device with SNMP v2

[15-Sep-2011 16:01:32] <rmatte> make sure the interface type shows up as ethernetCsmacd_64

[15-Sep-2011 16:02:22] <rmatte> SNMP v1 doesn't support 64bit counters which are required to monitor gig interfaces

[15-Sep-2011 16:03:03] <rmatte> Zenoss defaults to using SNMP v1 for some reason

[15-Sep-2011 16:03:08] <rmatte> it needs to be changed in zProperties

[15-Sep-2011 16:03:34] <rmatte> if it is v1 and you change it to v2 remodel your devices to pick up the interfaces as 64bit

[15-Sep-2011 16:04:01] <humanix> i see

[15-Sep-2011 16:04:04] <rmatte> you'll also have to recreate your RRDs, or at least increase the max value on them

[15-Sep-2011 16:04:07] <humanix> let me review it

[15-Sep-2011 16:04:10] <rmatte> k

[15-Sep-2011 16:04:40] <humanix> it seems to be that, im making the necesary changes

[15-Sep-2011 16:04:47] <rmatte> k

[15-Sep-2011 16:06:21] <humanix> do i lose historic statis if i do this?

[15-Sep-2011 16:06:49] <humanix> maybe if it is possible to convert data from ethernetcsmacd rrd to this one

[15-Sep-2011 16:14:57] <rmatte> humanix: I've done the conversion myself before but I had to script it out

[15-Sep-2011 16:15:34] <rmatte> if you can afford to lose your historical data that's the easiest way

[15-Sep-2011 16:15:41] <rmatte> otherwise you'll have to script out all the changes

[15-Sep-2011 16:16:07] <rmatte> need to copy the data over to the new location, rename it (the RRD files), then use RRD tool on each of them to increase the maximum on them

[15-Sep-2011 16:16:43] <rmatte> and it needs to be done fairly intelligently since not all interfaces are 64bit

[15-Sep-2011 16:17:40] <humanix> i see

[15-Sep-2011 16:17:50] <humanix> i already lose the data

[15-Sep-2011 16:17:56] <humanix> and i wanst supposed so

[15-Sep-2011 16:17:57] <rmatte> the data is technically garbage anyways

[15-Sep-2011 16:18:09] <humanix> so my boss is going to kill me

[15-Sep-2011 16:18:14] <humanix> cause they are doing capacity planning

[15-Sep-2011 16:18:14] <humanix> lol

[15-Sep-2011 16:18:23] <rmatte> you'll lose it because the new RRD files have _64 in the name

[15-Sep-2011 16:18:29] <rmatte> so it won't recognize the old ones

[15-Sep-2011 16:18:37] <rmatte> you haven't technically lost the data

[15-Sep-2011 16:18:40] <rmatte> it's still there

[15-Sep-2011 16:18:49] <humanix> can i revert to the old data?

[15-Sep-2011 16:18:55] <rmatte> what you have to do is find any interfaces that have the new RRDs with _64

[15-Sep-2011 16:19:00] <rmatte> and copy the old RRDs to the new name

[15-Sep-2011 16:19:19] <rmatte> then you need to use rrdtool against all of them to increase the maximum on them

[15-Sep-2011 16:19:33] <rmatte> hold on, I may have some code still laying around from when I did mine

[15-Sep-2011 16:19:39] <humanix> all ports on this machine are giga or 10 giga

[15-Sep-2011 16:20:03] <humanix> if i put snmpv1

[15-Sep-2011 16:20:10] <rmatte> I understand that, but it didn't pick them up as 64bit before because you were using snmpv1

[15-Sep-2011 16:20:17] <humanix> will it fall back to use old rrds?

[15-Sep-2011 16:20:33] <rmatte> if you put it back to v1 and remodel, yes, but why wouldn't you want to actually fix this?

[15-Sep-2011 16:20:45] <rmatte> you can't monitor gig interfaces with snmp v1, it just doesn't work

[15-Sep-2011 16:20:50] <rmatte> the counters hit max values and roll over

[15-Sep-2011 16:20:57] <rmatte> so the data your were collecting is all bad data anyways

[15-Sep-2011 16:21:02] <rmatte> inaccurate

[15-Sep-2011 16:21:16] <gwb235> ugh.  unless you have crappy old netapps that have a 32bitlow and 32bit high counter

[15-Sep-2011 16:21:28] <rmatte> well yeh

[15-Sep-2011 16:21:32] <rmatte> but that's not the case here

[15-Sep-2011 16:21:33] <rmatte> lol

[15-Sep-2011 16:21:36] <gwb235> glad i don't have to use that ugly glue code anymore

[15-Sep-2011 16:21:38] <rmatte> and that'd be a custom template too

[15-Sep-2011 16:22:35] <rmatte> hold on, I'll give you a way to get your data back

[15-Sep-2011 16:22:43] <rmatte> without having to re-enable v1

[15-Sep-2011 16:22:55] <rmatte> just need to make some slight adjustments to a script I wrote

[15-Sep-2011 16:23:21] <humanix> will be great

[15-Sep-2011 16:25:16] <rmatte> ok, so...

[15-Sep-2011 16:25:22] <rmatte> http://fpaste.org/xrhm/raw/

[15-Sep-2011 16:25:34] <rmatte> ssh to the server

[15-Sep-2011 16:25:38] <rmatte> become the zenoss user

[15-Sep-2011 16:25:43] <rmatte> save that script to a .sh file

[15-Sep-2011 16:25:47] <rmatte> chmod +x the file

[15-Sep-2011 16:25:49] <rmatte> the execute it

[15-Sep-2011 16:26:07] <nyeates> Ryan, on the graph refresh ticket

[15-Sep-2011 16:26:13] <nyeates> what browser did you test on?

[15-Sep-2011 16:26:14] <rmatte> that'll rename all of your old RRDs to 64bit RRDs (provided it finds a new 64bit RRD in the directory)

[15-Sep-2011 16:26:24] <rmatte> you may want to back up your RRD directory before you do this just in case

[15-Sep-2011 16:26:43] <rmatte> as zenoss user: cp -R $ZENHOME/perf ~/

[15-Sep-2011 16:26:54] <rmatte> then when you're done and happy you can delete ~/perf

[15-Sep-2011 16:27:03] <rmatte> nyeates: firefox

[15-Sep-2011 16:27:10] <nyeates> version?

[15-Sep-2011 16:27:16] <rmatte> let me check...

[15-Sep-2011 16:27:31] <humanix> rmate should i do changes in rrd anyway right?

[15-Sep-2011 16:27:33] <rmatte> 3.6.13

[15-Sep-2011 16:27:48] <rmatte> humanix: hunh?

[15-Sep-2011 16:28:10] <rmatte> humanix: what are you not understanding?

[15-Sep-2011 16:29:54] <humanix> i understand that the script are renaming the files in order to get zenoss graphs again

[15-Sep-2011 16:30:00] <rmatte> correct

[15-Sep-2011 16:30:03] <humanix> but you spoke me about the limit of rrds

[15-Sep-2011 16:30:14] <rmatte> oh, yeh, hold on, I have a script for that too

[15-Sep-2011 16:30:40] <humanix> but old data is garbage anyways right?

[15-Sep-2011 16:30:50] <rmatte> technically yes, but if you want to keep it you can

[15-Sep-2011 16:30:55] <humanix> i see

[15-Sep-2011 16:32:20] <rmatte> http://fpaste.org/YBJa/raw/

[15-Sep-2011 16:32:30] <rmatte> there's a script to increase the max value of the RRDs once you copy them

[15-Sep-2011 16:32:37] <rmatte> it increases it to 10gbit

[15-Sep-2011 16:33:07] <rmatte> you need to be in $ZENHOME/perf when you execute it

[15-Sep-2011 16:33:11] <rmatte> cd $ZENHOME/perf

[15-Sep-2011 16:33:13] <rmatte> then run the script

[15-Sep-2011 17:17:59] kocolosk is now known as kocolosk|away

[15-Sep-2011 17:18:35] <humanix> ok

[15-Sep-2011 17:18:36] <humanix> right

[15-Sep-2011 17:18:39] <humanix> thanks rmatte

[15-Sep-2011 17:30:18] <rmatte> This is the funniest thing I've seen in a while: http://www.youtube.com/watch?v=GI6CfKcMhjY&feature=share

[15-Sep-2011 17:30:34] <rmatte> humanix: no problem

[15-Sep-2011 20:53:35] JuStIciA is now known as JuStIcIa_

[15-Sep-2011 21:27:51] kocolosk|away is now known as kocolosk

[15-Sep-2011 22:57:47] kocolosk is now known as kocolosk|away

[16-Sep-2011 00:00:01] [disconnected at Fri Sep 16 00:00:01 2011]

[16-Sep-2011 00:00:01] [connected at Fri Sep 16 00:00:01 2011]

[16-Sep-2011 00:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[16-Sep-2011 06:38:37] <enkrypt> hi guys. I upgraded yesterday from 3.1.0 to 3.2.0 and some of my OSProcess monitoring broke. I used to be monitoring 9 processes, called drbd0_asender, drbd0_receiver, drbd0_worker and so on. since version 3.2.0 that stopped working, he reports all of them as failed, except for ONE of them, and that one has a process count of 9 according to zenoss. is this a known issue?

[16-Sep-2011 07:24:18] <enkrypt> are there any known bugs about not being able to connect to ZenHub, in the case of distributed collectors?

[16-Sep-2011 07:24:25] <enkrypt> I started seeing this error since upgrading to 3.2.0

[16-Sep-2011 07:39:39] <enkrypt> I've also got these errors since 3.2.0:

[16-Sep-2011 07:39:40] <enkrypt> WARNING zen.zenucsevents: No service EventService named: ZenHub may be disconnected

[16-Sep-2011 07:39:45] <enkrypt> ERROR zen.zenucsevents: Maintenance failed: ZenHub is down

[16-Sep-2011 07:39:51] <whyzgeek> hi

[16-Sep-2011 07:56:44] <enkrypt> I keep getting messages from my remote collector saying that zenhub is down. but zenhub is up, checked myself connectivity with telnet. and with tcpdump on port 8789 on the hub I see incoming + outgoing traffic. is this a 3.2.0 bug?

[16-Sep-2011 09:05:11] <Hackman238> Morning All

[16-Sep-2011 09:09:30] <tsener> morning

[16-Sep-2011 09:10:51] <jmp242> morning

[16-Sep-2011 09:28:41] <jmp242> Hey, anyone want to look at this forum thread? I haven't used 3.2 yet (waiting for the 3.2.1) but: message/61408#61408

[16-Sep-2011 09:29:19] <dhopp> jmp242:  looking now

[16-Sep-2011 09:30:36] <dhopp> I think he's just asking how to do it in 3.x..it's not really a 3.2 specific question

[16-Sep-2011 09:31:13] <jmp242> Oh, ok

[16-Sep-2011 09:31:17] <jmp242> I misunderstood then

[16-Sep-2011 09:31:18] <dhopp> I'm responding

[16-Sep-2011 09:31:21] <jmp242> ok

[16-Sep-2011 09:31:37] <Jane_Curry> Afternoon....

[16-Sep-2011 09:32:14] <jmp242> Hi Jane_Curry

[16-Sep-2011 09:32:34] <jmp242> still earlyish morning here

[16-Sep-2011 09:33:09] <Jane_Curry> and half way to wine time here

[16-Sep-2011 09:33:36] <Jane_Curry> dhopp - you got the forum thread sussed?  I answered this one for a client only yesterday so can do it fast if you like

[16-Sep-2011 09:36:14] <Jane_Curry> jmp242: I'm a bit confused about when ZCA meet Zenoss - today? next Friday? both?

[16-Sep-2011 09:36:49] <dhopp> Jane_Curry:  I already answered it :-)

[16-Sep-2011 09:38:04] <Jane_Curry> dhopp - do you have a good answer to this one though.....

[16-Sep-2011 09:38:39] <dhopp> Jane_Curry:  wait..I'm confused...

[16-Sep-2011 09:38:46] <Jane_Curry> If you want to create  a copy of a component template for all that component on a device - how do you do it??

[16-Sep-2011 09:38:46] <dhopp> are you talking about the one jmp242 referenced?

[16-Sep-2011 09:39:22] <dhopp> Jane_Curry:  oh..hrm..I'm not sure how you would do that..you would have to do it one by one or do something in the dmd I would imagine

[16-Sep-2011 09:39:35] <Jane_Curry> If interface is your component example, your answer works per interface - but if you have 200 interfaces you are not going to change each one

[16-Sep-2011 09:39:57] <dhopp> Jane_Curry:  true..but he was asking for just one filesystem I thought

[16-Sep-2011 09:40:21] <dhopp> because I have servers where /usr/local should have a different threshold then /var/lib/mysql for example

[16-Sep-2011 09:40:23] <Jane_Curry> I ended up creating a device subclass (as my guy had 3 devices where he wanted a component threshold changed), copied the template to the new dev class

[16-Sep-2011 09:40:43] <Jane_Curry> and then copied the 3 devices to that subclass - messy though

[16-Sep-2011 09:41:26] <Jane_Curry> dhopp - yeh - the forum append WAS asking for 1 filesystem and your answer is dead right for him

[16-Sep-2011 09:42:22] <Jane_Curry> Any devs around????

[16-Sep-2011 09:42:51] kocolosk|away is now known as kocolosk

[16-Sep-2011 09:43:08] <dhopp> Jane_Curry:  and you'll notice I did say "there are a couple of different ways to do it"…hehe :-)

[16-Sep-2011 09:57:47] <jmp242> Jane_Curry: we thought it was today, but they moved it to next friday

[16-Sep-2011 09:59:39] <Jane_Curry> jmp242 - so today is just ZCA - no update from Zenoss (like on documentation progress) until next Friday?

[16-Sep-2011 10:10:04] <Hackman238> Jane_Curry: That was my mistake

[16-Sep-2011 10:10:16] <Hackman238> Jane_Curry: I have some unofficial words they will officiate next meeting

[16-Sep-2011 10:13:59] <Jane_Curry> huh - do you mean you can share words now or no-one gets anything until next week?

[16-Sep-2011 10:16:20] <jmp242> Jane_Curry: I'm sure it cannot be shared in open IRC

[16-Sep-2011 10:17:56] <Hackman238> Jane_Curry: I'll tell you now, they'll make it official work next week

[16-Sep-2011 10:18:06] <Hackman238> *word

[16-Sep-2011 10:35:28] <rocket> Jane_Curry: there is a defect in for your component problem, what you did is probably the best current work around

[16-Sep-2011 10:41:03] <photon01> Hi every1

[16-Sep-2011 10:41:16] <photon01> I need some help with the Zope daemon plz :s

[16-Sep-2011 10:43:23] <Jane_Curry> <rocket> - sorry - not sure what you are referring to here???

[16-Sep-2011 10:44:02] <rocket> different component templates for interfaces

[16-Sep-2011 10:44:26] <rocket> 08:38 <@Jane_Curry> If you want to create  a copy of a component template for all that component  on a device - how do you do it??

[16-Sep-2011 10:46:07] <Jane_Curry> Rocket - gotcha - do you have the ticket number??

[16-Sep-2011 10:48:54] <rocket> dont recall from memory

[16-Sep-2011 10:52:07] <rocket> externally 7579

[16-Sep-2011 10:52:26] <photon01> any1 can help me with the zope daemon ?

[16-Sep-2011 10:57:46] <chemist> hello all

[16-Sep-2011 10:59:38] <jmp242> photon01: wait till noonish

[16-Sep-2011 10:59:40] <jmp242> EST

[16-Sep-2011 11:00:12] <Jane_Curry> rocket: doesn't look like 7579 is going anywhere though - http://dev.zenoss.com/trac/ticket/7579

[16-Sep-2011 11:00:16] <photon01> <jmp242> : ok thanks

[16-Sep-2011 11:01:00] <rocket> I just said it was reported .. its up to engineering to get enough cycles to address it

[16-Sep-2011 11:01:36] <chemist> can  upgrade from 3.0.3 to 3.20 with the stack installer?

[16-Sep-2011 11:06:26] <jmp242> chemist: you need to go to 3.1 first

[16-Sep-2011 11:06:28] <jmp242> then 3.2

[16-Sep-2011 11:06:39] <jmp242> but you need to patch 3.2 also - see room title

[16-Sep-2011 11:06:39] <chemist> jmp242: ok thanks

[16-Sep-2011 11:06:56] <chemist> need to apply all those patches?

[16-Sep-2011 11:07:22] <chemist> might just go to 3.1 for now

[16-Sep-2011 11:07:26] <chemist>

[16-Sep-2011 11:07:41] <Jane_Curry> chemist - 3.1 might be a good solution for now

[16-Sep-2011 11:11:18] <rocket> if you are having any modelling performance issues going to 3.2 is still better even with the need to patch them

[16-Sep-2011 11:16:49] <photon01> (sorry for the spam) any1 can help with a zope daemon problem ?

[16-Sep-2011 11:30:49] <enkrypt> is there any known issue introduced with zenoss 3.2.0 related to zenprocess? I've got several processes monitored that used to work fine on 3.1.0, but not anymore on 3.2.0

[16-Sep-2011 11:32:48] <rocket> enkrypt: applied the patches?

[16-Sep-2011 11:33:02] <enkrypt> rocket: yes, all 3 of them

[16-Sep-2011 11:33:19] <rocket> then I do not believe there are any currently known issues

[16-Sep-2011 11:33:34] <rocket> are there tracebacks in your logs that might give clues?

[16-Sep-2011 11:34:57] <enkrypt> rocket: unfortunately no

[16-Sep-2011 11:35:08] <enkrypt> 2011-09-16 17:34:42,495 INFO zen.zenprocess: Searching for possible matches for set([drbd1_worker, drbd0_receiver, drbd2_worker, drbd0_worker, drbd0_asender, drbd2_receiver, drbd1_asender, drbd1_receiver])

[16-Sep-2011 11:35:12] <enkrypt> 2011-09-16 17:34:42,506 WARNING zen.zenprocess: (193.58.14.66) Process not running: drbd0_asender  Using regex '\Adrbd\d+_'

[16-Sep-2011 11:35:28] <enkrypt> that is the message I get. and that process is running with 100% certainty

[16-Sep-2011 11:36:18] <enkrypt> also I get periodical halts of zenhub I think

[16-Sep-2011 11:36:26] <enkrypt> the UI sometimes claims that he can't contact the server

[16-Sep-2011 11:36:31] <enkrypt> and a few seconds later it's ok

[16-Sep-2011 11:36:42] <enkrypt> also in the logs on my remote collector I sometimes can't connec to zenhub

[16-Sep-2011 11:38:41] <enkrypt> and there you have it, zenperfsnmp stopped working since 10 minutes

[16-Sep-2011 11:38:51] <enkrypt> I'm not sure what is going on, something is very unstable since upgrading to 3.2.0

[16-Sep-2011 11:40:08] <enkrypt> rocket: the main issue I think is with zenhub

[16-Sep-2011 11:40:25] <enkrypt> rocket: my remote collector often gives an error saying he can't connect to zenhub

[16-Sep-2011 11:40:35] <enkrypt> rocket: and the GUI often simply doesn't work

[16-Sep-2011 11:41:45] <rocket> zenhub have any workers assigned?

[16-Sep-2011 11:42:15] <enkrypt> yes 7 workers

[16-Sep-2011 11:42:17] <enkrypt> I see this now:

[16-Sep-2011 11:42:21] <rocket> sounds mostly like a tuning issue .. if there arent enough workers there is no chance for the daemons to connect.

[16-Sep-2011 11:42:22] <enkrypt> 2011-09-16 17:41:22,026 DEBUG zen.ZenHub: all workers are busy

[16-Sep-2011 11:43:17] <rocket> enkrypt: are you an enterprise customer?

[16-Sep-2011 11:43:23] <enkrypt> so any idea why 7 workers was fine before the upgrade, but after 3.2.0 it isn't? can I just increase this number? I've got 8 physical CPU cores (16 if you count the hyperthreading story)

[16-Sep-2011 11:43:33] <enkrypt> rocket: yes, name is Jan Geboers, working for Cegeka Belgium

[16-Sep-2011 11:44:05] <rocket> Jan you can increase the workers .. I would not know what may have changed other than extra modelling after a restart

[16-Sep-2011 11:44:27] <rocket> I would open a ticket so we can track down what is consuming worker time.

[16-Sep-2011 11:44:37] <rocket> if I had to guess it would be modelling right now

[16-Sep-2011 11:45:54] <enkrypt> ok thanks, will increase them a bit.

[16-Sep-2011 11:47:53] <nexex> Is there a way to disable all monitoring for a device, temporarily?

[16-Sep-2011 11:48:05] <rmatte> yes, move it to Decommissioned production state\

[16-Sep-2011 11:48:07] <rmatte> state*

[16-Sep-2011 11:48:43] <rmatte> that will halt monitoring of it

[16-Sep-2011 11:48:44] <hmp> blah, NOC job is sooo boring

[16-Sep-2011 11:49:01] <rmatte> hmp: depends on what you're actually doing at the NOC

[16-Sep-2011 11:49:11] <rmatte> if you're an analyst, then yeh, probably

[16-Sep-2011 11:49:50] <hmp> rmatte: we have cyles where one week in every eight you just sit in a monitoring room with 2 giant plasma screens and monitor the network

[16-Sep-2011 11:50:20] <hmp> and serve as higher tier support to helpdesk

[16-Sep-2011 11:50:59] <hmp> 7 other weeks is just fine, developing inhouse monitoring systems, developing the network, etc.

[16-Sep-2011 11:51:02] <Hackman238> hmp: Should not be a help desk

[16-Sep-2011 11:51:10] <Hackman238> hmp: LOL

[16-Sep-2011 11:51:11] <rmatte> hmp: that's such a stupid way to do it, we have our monitoring tools directly integrated with our ticketing system, so tickets get opened automatically and the analysts just have to work them

[16-Sep-2011 11:51:23] <rmatte> hmp: that's how they did it at a past company that I worked at so that's how we set it up here

[16-Sep-2011 11:51:25] <Hackman238> rmatte: Win, thats what we do

[16-Sep-2011 11:51:45] <hmp> rmatte: we too have automation that opens tickets

[16-Sep-2011 11:51:53] <Hackman238> rmatte: (meaning Win for your automation)

[16-Sep-2011 11:51:54] <hmp> so you could say i'm an analyst this week

[16-Sep-2011 11:52:05] <Hackman238> hmp: Darn

[16-Sep-2011 11:52:11] <photon01> any1 can help me with a zope daemon problem ?

[16-Sep-2011 11:52:26] <rmatte> enkrypt: that's completely normal, it needs to gather data for 3 cycles (15 minutes) before it can display the memory info, unless you're trying to monitor Windows 2000 boxes which don't present memory information via SNMP

[16-Sep-2011 11:52:32] <Hackman238> photon01: Whats up?

[16-Sep-2011 11:52:34] <rmatte> erm, was meant for privmsg

[16-Sep-2011 11:52:54] <Hackman238> rmatte: Now we know all your secrets! HAHAHA

[16-Sep-2011 11:52:58] <Hackman238> rmatte: JK

[16-Sep-2011 11:52:59] <rmatte> lol

[16-Sep-2011 11:53:10] <hmp> Hackman238: yea, but we also serve if one of the memebers of our NREN calls helpdesk and asks for support

[16-Sep-2011 11:53:20] <hmp> then they redirect it to us, couse they dont know shit

[16-Sep-2011 11:53:28] <photon01> <Hackman238> : When I start Zope daemon it shows daemon manager not running

[16-Sep-2011 11:53:37] <Hackman238> hmp: LOL

[16-Sep-2011 11:53:42] <Hackman238> photon01: Gotcha

[16-Sep-2011 11:53:46] <Hackman238> photon01: su zenoss

[16-Sep-2011 11:53:51] <Hackman238> photon01: zenoss stop

[16-Sep-2011 11:53:51] <hmp> Hackman238: its not funny

[16-Sep-2011 11:53:58] <Hackman238> photon01: zeoctl {enter}

[16-Sep-2011 11:54:04] <Hackman238> photon01: fg {enter}

[16-Sep-2011 11:54:11] <Hackman238> photon01: Does any trace or errors show?

[16-Sep-2011 11:54:45] <hmp> btw, is there a zenpack to poll interfaces?

[16-Sep-2011 11:54:47] <photon01> <Hackman238> : ok let me try (actually on Zenoss UI logs, it's zope daemon that's talkative, zeo seems quite)

[16-Sep-2011 11:55:16] <Hackman238> photon01: Yeah but I'm looking starting at the root

[16-Sep-2011 11:55:37] <photon01> <Hackman238> : okaay. well here's the output

[16-Sep-2011 11:55:49] <photon01> program: /usr/local/zenoss/zenoss/bin/runzeo

[16-Sep-2011 11:55:49] <photon01> daemon manager not running

[16-Sep-2011 11:55:49] <photon01> zeoctl> fg

[16-Sep-2011 11:55:49] <photon01> /usr/local/zenoss/zenoss/bin/runzeo

[16-Sep-2011 11:56:42] <photon01> <Hackman238> : I think it's because I switched data.fs and Data.fs.old once

[16-Sep-2011 11:56:55] <photon01> <Hackman238> : but I switched them back again

[16-Sep-2011 11:57:05] <Hackman238> photon01: Okay, looks good

[16-Sep-2011 11:57:11] <Hackman238> photon01: CTRL + D to exit

[16-Sep-2011 11:57:21] <Hackman238> photon01: do a ps aux | grep zeo

[16-Sep-2011 11:57:26] <Hackman238> photon01: Make sure zeo stopped

[16-Sep-2011 11:57:34] <photon01> ctrl +D doesn't cork

[16-Sep-2011 11:57:34] <Hackman238> photon01: If yes, zeoctl start; zopectl fg

[16-Sep-2011 11:57:38] <photon01> ctrl c ?

[16-Sep-2011 11:57:44] <photon01> work

[16-Sep-2011 11:57:51] <Hackman238> photon01: My bad, sorry

[16-Sep-2011 11:57:56] <photon01> okay ^^

[16-Sep-2011 11:58:23] <photon01> zeo is stopped

[16-Sep-2011 11:58:54] <photon01> :-o

[16-Sep-2011 11:58:57] <Hackman238> photon01: Okay, zeoctl start; zopectl fg

[16-Sep-2011 11:58:58] <photon01> A python exception

[16-Sep-2011 11:59:12] <photon01> ZConfig.ConfigurationError: There was a problem starting a server of type "HTTPServer". This may mean that your user does not have permission to bind to the port which the server is trying to use or the port may already be in use by another application. (Address already in use)

[16-Sep-2011 11:59:21] <Hackman238> photon01: Ah nice

[16-Sep-2011 11:59:26] <rmatte> If Zenoss aren't going to release 3.2.1 they should seriously consider pulling back on the 3.2 release, people are reporting other issues with it

[16-Sep-2011 11:59:29] <Hackman238> photon01: do a ps aux | grep zope

[16-Sep-2011 11:59:47] <Hackman238> kill them all, if they wont die, try kill -7, last resort -9

[16-Sep-2011 11:59:56] <photon01> okay

[16-Sep-2011 12:00:01] [disconnected at Fri Sep 16 12:00:01 2011]

[16-Sep-2011 12:00:01] [connected at Fri Sep 16 12:00:01 2011]

[16-Sep-2011 12:00:12] [zenoss-logger (logger bot) has joined #zenoss]

[16-Sep-2011 12:00:18] <Hackman238> photon01: Squid isnt running on port 8080 by chance, right?

[16-Sep-2011 12:00:25] <photon01> no squid there

[16-Sep-2011 12:00:38] <Hackman238> photon01: su zenoss; zenoss stop; zenoss start

[16-Sep-2011 12:00:48] <Hackman238> photon01: Try to hit the web UI

[16-Sep-2011 12:00:58] <photon01> okay

[16-Sep-2011 12:01:42] <photon01> actually all worked fine, except when I tried installing a zenpack. It kept in broken status

[16-Sep-2011 12:01:57] <photon01> (even though its templates seemed available)

[16-Sep-2011 12:02:01] <Hackman238> photon01: A zenpack you installed still says 'broken' under zenpacks?

[16-Sep-2011 12:02:24] <photon01> yes, but the issue came before that

[16-Sep-2011 12:02:35] <photon01> UI  works fine

[16-Sep-2011 12:03:04] <photon01> ah ! it's not broken anymore !

[16-Sep-2011 12:03:09] <Hackman238> photon01:

[16-Sep-2011 12:03:12] <photon01>

[16-Sep-2011 12:03:20] <photon01> we did nothing to resolve the issue, no ?

[16-Sep-2011 12:03:49] <rocket> rmatte: 3.2.1 is being built there should be a 3.2.1 beta shortly as far as I know

[16-Sep-2011 12:04:00] <rmatte> rocket: good to hear

[16-Sep-2011 12:04:10] <Hackman238> photon01: You sometimes need to restart zope to get zenpacks to register

[16-Sep-2011 12:05:17] <photon01> <Hackman238> : I did it before

[16-Sep-2011 12:05:49] <photon01> <Hackman238> : what I dont understand is why the "daemon manager not running" error doesnt show up anymore

[16-Sep-2011 12:06:11] <Hackman238> photon01: The pid file wasnt in $ZENHOME/var/

[16-Sep-2011 12:06:30] <Hackman238> photon01: restarting zope after killing the remaining daemons recreated it

[16-Sep-2011 12:07:20] <photon01> <Hackman238> : okay ! still not that clear for me

[16-Sep-2011 12:07:31] <photon01> <Hackman238> : thanks a lot ! that was my prod environment

[16-Sep-2011 12:07:43] <Hackman238> photon01: No problem, anytime!

[16-Sep-2011 13:56:41] <JohnnyNOC> Hackman238 you around?

[16-Sep-2011 14:00:34] <Hackman238> JohnnyNOC: Hey, how goes it?

[16-Sep-2011 14:03:44] <JohnnyNOC> going well

[16-Sep-2011 14:04:06] <JohnnyNOC> trying to figure out why ssSwapIn and ssSwapOut OIDs on CentOS machines don't seem to returning the correct info

[16-Sep-2011 14:04:17] <JohnnyNOC> we had a machine swapping but zenoss didn't report it

[16-Sep-2011 14:04:35] <JohnnyNOC> wondering fi you had any ideas why

[16-Sep-2011 14:04:56] <JohnnyNOC> i'm thinking we might need a kernel upgrade or maybe just upgrade net-snmp but i don't have anything concrete to indicate that's going to be our fix

[16-Sep-2011 14:05:56] <Hackman238> JohnnyNOC: If you pull the OID's usng snmp-get, are sane values provided?

[16-Sep-2011 14:07:57] <JohnnyNOC> yes

[16-Sep-2011 14:08:23] <JohnnyNOC> but apparently incorrect because sar said it was swapping during a certain period but zenoss isn't showing it

[16-Sep-2011 14:09:40] <Hackman238> JohnnyNOC: Sounds like its probably a netsnmp bug

[16-Sep-2011 14:09:52] <Hackman238> JohnnyNOC: Does the version match other boxes operating normally?

[16-Sep-2011 14:10:50] <JohnnyNOC> it's a little older

[16-Sep-2011 14:11:01] <JohnnyNOC> 5.4.3 vs 5.3.2

[16-Sep-2011 14:11:04] <Hackman238> JohnnyNOC: Try updating net-snmp

[16-Sep-2011 14:11:14] <Hackman238> JohnnyNOC: then remodel for good measure

[16-Sep-2011 14:11:58] <JohnnyNOC> *nod*

[16-Sep-2011 14:46:44] <Hackman238> JohnnyNOC: Did that work?

[16-Sep-2011 14:55:33] <rmatte> strange, I run the aggregate network graph and I see a spike that's measured in petabytes, yet when I run a standard interface report for the period I see nothing even close that

[16-Sep-2011 14:55:47] <rmatte> wonder how I can track down this garbage data that's causing it

[16-Sep-2011 14:57:29] <Hackman238> rmatte: I have a script that clips peaks automagically

[16-Sep-2011 14:57:34] <Hackman238> rmatte: You want it?

[16-Sep-2011 14:57:58] <rmatte> how does it do it?

[16-Sep-2011 14:58:07] <rmatte> dumps to xml, clips, restores?

[16-Sep-2011 14:58:10] <rmatte> or some other way?

[16-Sep-2011 14:58:28] <rmatte> and yes, I'd appreciate it if you could send it to me

[16-Sep-2011 14:59:15] <JohnnyNOC> Hackman238 unfortunately newer net-snmp doesn't help

[16-Sep-2011 14:59:16] <JohnnyNOC> :/

[16-Sep-2011 14:59:21] <rmatte> I wrote a script that clips negative values from RRDs... I could probably hack something together to do it for this too, but if you already have something I might as well use that

[16-Sep-2011 15:00:36] <Hackman238> rmatte: Sent

[16-Sep-2011 15:00:49] <Hackman238> rmatte: dumps, reads, figured, clips, restores

[16-Sep-2011 15:01:04] <Hackman238> JohnnyNOC: Darn- thats odd

[16-Sep-2011 15:01:14] <JohnnyNOC> i know, boo

[16-Sep-2011 15:01:18] <Hackman238> JohnnyNOC: Has this box been rebooted in a long while?

[16-Sep-2011 15:01:33] <JohnnyNOC> i ran something called memeater and watched it eat up to 66% of my swap but ssswapin and ssswapout oids still says 0 kb

[16-Sep-2011 15:01:42] <JohnnyNOC> this is a brand new box i'm testing on

[16-Sep-2011 15:01:44] <Hackman238> JohnnyNOC: Generally something simple like snmpgets for template based thresholding is all or nothing, not hit and miss

[16-Sep-2011 15:01:47] <JohnnyNOC> tho its running centos 5.5

[16-Sep-2011 15:02:17] <Hackman238> JohnnyNOC: Ah good test

[16-Sep-2011 15:02:19] <JohnnyNOC> i hear you.. obviously something's amiss..  i can see in /proc/meminfo my swap used as well, but still nothing in the OIDs

[16-Sep-2011 15:02:29] <jmp242> ZCA Update everyone: community/zca/blog/2011/09/16/zca-update-for-september-16th-2011--operations-roadmap-update

[16-Sep-2011 15:02:30] <JohnnyNOC> just 0 kb

[16-Sep-2011 15:02:31] <JohnnyNOC> boo

[16-Sep-2011 15:04:39] <Hackman238> JohnnyNOC: Thats strange. A Quick google doesnt render any obvious hits.

[16-Sep-2011 15:05:01] <Hackman238> JohnnyNOC: Try removing snmpd and related libraries, removing /etc/smnpd/snmpd.conf and reinstalling it

[16-Sep-2011 15:05:06] <JohnnyNOC> i noticed!  heheh..

[16-Sep-2011 15:05:16] <JohnnyNOC> pretty much did that when i wiped out 5.3.2 and installed 5.4.3

[16-Sep-2011 15:05:17] <Hackman238> JohnnyNOC: It could be a weird SElinux thing too, if thats not set to disabled

[16-Sep-2011 15:05:41] <JohnnyNOC> disabled

[16-Sep-2011 15:05:41] <JohnnyNOC> :[

[16-Sep-2011 15:06:47] <Hackman238> JohnnyNOC: Very odd

[16-Sep-2011 15:06:55] <JohnnyNOC> posting to net-snmp-users

[16-Sep-2011 15:06:59] <JohnnyNOC> hopefully someone knwos something

[16-Sep-2011 15:07:13] <Hackman238> JohnnyNOC: Good idea

[16-Sep-2011 15:09:28] <JohnnyNOC> while you're here i do have another qeustino though

[16-Sep-2011 15:09:49] <JohnnyNOC> i've got some events that aren't auto-clearing from my dashboard and can't seem to understand why

[16-Sep-2011 15:10:12] <Hackman238> http://www.foxnews.com/us/2011/09/16/denver-men-accused-taking-friends-corpse-on-boys-night-out/

[16-Sep-2011 15:10:36] <Hackman238> JohnnyNOC: It could be the clear event was part of an event queue that was truncated

[16-Sep-2011 15:10:44] <JohnnyNOC> hrm

[16-Sep-2011 15:10:57] <JohnnyNOC> let me take a look at some graphs but i don't think that's the issue

[16-Sep-2011 15:11:03] <JohnnyNOC> i dont think my queues are getting that high and are being timmed

[16-Sep-2011 15:11:05] <JohnnyNOC> trimmed

[16-Sep-2011 15:11:25] <Hackman238> JohnnyNOC: Also could be the clear was mis identified as a 'useless' clear event and disgarded

[16-Sep-2011 15:13:14] <JohnnyNOC> how do i figure out if that's what's going on?

[16-Sep-2011 15:13:30] <JohnnyNOC> my event queues only occasionally hit high numbers..  like 2-3x day tops, otherwise they're all fairly low

[16-Sep-2011 15:13:33] <rmatte> Hackman238: yeh, your script works like the one I wrote for removing negative values, nice

[16-Sep-2011 15:13:47] <jmp242> Hackman238: can you update the topic?

[16-Sep-2011 15:14:55] <Hackman238> Sure

[16-Sep-2011 15:15:18] <rmatte> Hackman238: ok, I need to understand... "Where: number is the % limit of spikes to chop"

[16-Sep-2011 15:15:21] <rmatte> %?

[16-Sep-2011 15:15:34] <rmatte> is it a literal number to chop above, or a percentage?

[16-Sep-2011 15:15:49] <Hackman238> %

[16-Sep-2011 15:16:03] <Hackman238> Backup your files first also

[16-Sep-2011 15:16:14] <rmatte> I don't get that though, a percentage of what?

[16-Sep-2011 15:16:30] <Hackman238> rmatte: Of a period

[16-Sep-2011 15:16:38] <rmatte> I basically want to chop any value above 10000000000

[16-Sep-2011 15:16:46] <Hackman238> rmatte: Heres how we use it: for i in `find os/interfaces/ -name ifHC*.rrd`; do /home/zenoss/scripts/removespikes.pl $i; done;

[16-Sep-2011 15:17:15] <Hackman238> rmatte: Of it doesnt do that

[16-Sep-2011 15:17:37] <rmatte> ok, I think I'll just mod my script to do it

[16-Sep-2011 15:17:48] <Hackman238> rmatte: What it does is find the top 0.6% percentile, always insane spikes.

[16-Sep-2011 15:18:16] <rmatte> oh

[16-Sep-2011 15:18:18] <Hackman238> rmatte: We use that when we see petabyte and terabyte peaks

[16-Sep-2011 15:18:44] <rmatte> so the top percentile out of all interfaces you mean

[16-Sep-2011 15:19:01] <rmatte> or just in each file?

[16-Sep-2011 15:19:03] <Hackman238> out of all periods in the rra in the ds

[16-Sep-2011 15:19:11] <rmatte> ah

[16-Sep-2011 15:19:23] <rmatte> alright, I'll backup my RRDs and give it a go

[16-Sep-2011 15:19:40] <Hackman238> rmatte: No worries, we use it all the time. Backups are always encouraged though!

[16-Sep-2011 15:19:47] <Hackman238> rmatte: Exactly

[16-Sep-2011 15:20:02] <rmatte> so you don't actually have to specify a percentage it defaults to 0.6

[16-Sep-2011 15:20:12] <Hackman238> rmatte: correct

[16-Sep-2011 15:20:13] <rmatte> I always do backups before I do this sort of thing

[16-Sep-2011 15:20:13] <rmatte> lol

[16-Sep-2011 15:20:20] <Hackman238> rmatte: you can tweak it to be smaller if you like

[16-Sep-2011 15:20:30] <Hackman238> rmatte: we've found 0.4-0.6 seems the best

[16-Sep-2011 15:21:02] <Hackman238> We're up to 0.6 because We've so many providers and we're often DDoS targets...like tens of Gb usually

[16-Sep-2011 15:21:38] <Hackman238> ...so when FCP's and NOC folks redirect traffic around, counter spikes happen often

[16-Sep-2011 15:23:15] <Hackman238> No comments on the link I posted?

[16-Sep-2011 15:24:25] <rmatte> sorry, I'm bouncing around from screen to screen, busy day, multitasking

[16-Sep-2011 15:25:28] <Hackman238> rmatte: Ah np.

[16-Sep-2011 15:25:32] bender183 is now known as fridget

[16-Sep-2011 15:25:50] <rmatte> this spike thing is bad because it's keeping a report from going out lol

[16-Sep-2011 15:26:09] <rmatte> usually I can find these spikes just by running a regular interface report with MAXIMUM for the period

[16-Sep-2011 15:26:12] <rmatte> but not this time apparently

[16-Sep-2011 15:26:55] <Hackman238> rmatte: Ah gotcha. I hope my script can help

[16-Sep-2011 15:27:16] <Hackman238> Oh, BTW all, I've a really fast python RRD merging script now

[16-Sep-2011 15:27:35] <rmatte> merging?

[16-Sep-2011 15:27:43] <Hackman238> rmatte: Concatanates two RRD's

[16-Sep-2011 15:27:51] <rmatte> interesting

[16-Sep-2011 15:28:03] <rmatte> what's the core application of that?

[16-Sep-2011 15:28:43] <rmatte> this server has a bazillion RRDs, taking forever to copy

[16-Sep-2011 15:28:53] * rmatte taps his foot

[16-Sep-2011 15:29:18] <Hackman238> rmatte: Well we swap devices often. Ex. Converting Cores from 6509 to ASR 9K. This means the device is in the same place and does the same core function, but its ints are diffrent. We create a map of what maps to what and the script merges the old data with the new so we have complete graphs!

[16-Sep-2011 15:29:42] <rmatte> ah, we're doing a lot of conversions from 6509 to other devices for clients (mostly nexus)

[16-Sep-2011 15:29:55] <rmatte> I found another massively unimpressive bug with nexus today

[16-Sep-2011 15:30:11] <rmatte> when you poll CPU values, the only value it'll present are 0, 33, 50, 66, and 100

[16-Sep-2011 15:30:25] <rmatte> so if the value is actually 67 and above you end up with it always showing 100

[16-Sep-2011 15:30:31] <rmatte> stupidest thing I've seen in a while

[16-Sep-2011 15:30:42] fridget is now known as bender183

[16-Sep-2011 15:31:20] <Hackman238> rmatte: What!???

[16-Sep-2011 15:31:25] <Hackman238> rmatte: Thats insane

[16-Sep-2011 15:31:28] <rmatte> yup

[16-Sep-2011 15:31:37] <rmatte> we're having to disable CPU thresholds for Nexus devices because of that

[16-Sep-2011 15:31:42] <rmatte> false positives constantly

[16-Sep-2011 15:31:45] <Hackman238> rmatte: Madness!

[16-Sep-2011 15:32:04] <rmatte> Hackman238: ok, so the command you gave me is if you actually know what device you're doing it against

[16-Sep-2011 15:32:05] <rmatte> I don't

[16-Sep-2011 15:32:10] <rmatte> should this be safe to run globally?

[16-Sep-2011 15:32:36] <rmatte> keep in mind that all of my RRDs for standard and 64bit are named the same, I removed the HC

[16-Sep-2011 15:32:45] <rmatte> (for a few reasons that I won't get in to)

[16-Sep-2011 15:33:44] <rmatte> actually, I do have a few devices that I'll run it against first since I have a feeling that one of them may be the culprit (the nexus devices)

[16-Sep-2011 15:34:58] <Hackman238> rmatte: rmatte Yep

[16-Sep-2011 15:35:04] <Hackman238> Use the foo I gave you

[16-Sep-2011 15:35:24] <Hackman238> rmatte: Su zenoss

[16-Sep-2011 15:35:27] <Hackman238> for i in `find os/interfaces/ -name ifHC*.rrd`; do  /home/zenoss/scripts/removespikes.pl $i; done;

[16-Sep-2011 15:35:55] <Hackman238> If you want to traverse all devices, just change the find path

[16-Sep-2011 15:36:34] <rmatte> k

[16-Sep-2011 15:36:48] <rmatte> this shouldn't destroy legitimate data should it?

[16-Sep-2011 15:37:14] <Hackman238> rmatte: It shouldnt. Never has for us. Backup backup though

[16-Sep-2011 15:37:18] <rmatte> I mean, it's only removing one datasource per file right?

[16-Sep-2011 15:37:24] <Hackman238> rmatte: We have backups on spindle every 15 min

[16-Sep-2011 15:38:07] <Hackman238> rmatte: It wont remove it- it'll dump, process 0.6 percentile, calculate new values, insert in place of values to clip, restore

[16-Sep-2011 15:46:12] <rmatte> ah

[16-Sep-2011 15:46:14] <rmatte> ok cool

[16-Sep-2011 15:46:19] <rmatte> running it now...

[16-Sep-2011 15:51:04] <rmatte> Hackman238: that command doesn't seem to be handling filenames with spaces in them at all

[16-Sep-2011 15:51:17] <rmatte> I've tried adding quotes around the $i and even \"\"

[16-Sep-2011 15:51:20] <rmatte> neither works

[16-Sep-2011 15:51:50] <rmatte> maybe it's the script itself that's not handling them properly?

[16-Sep-2011 15:52:03] <rmatte> I'm not a huge perl guy so not sure

[16-Sep-2011 15:53:24] <Hackman238> rmatte: Hum...not sure. I never put spaces in datasources. Try running the script against an rrd in a path which includes spaces directly. That'll tell me if its the find or the script

[16-Sep-2011 15:53:43] <rmatte> it's not spaces in the datasource names

[16-Sep-2011 15:53:50] <rmatte> it's spaces in the interface names that it's complaining about

[16-Sep-2011 15:53:55] <rmatte> for the directories that the data is stored in

[16-Sep-2011 15:54:13] <Hackman238> Oh gotcha. Thats bizarre- let me check here

[16-Sep-2011 15:54:24] <Hackman238> Oh- nvm.

[16-Sep-2011 15:54:27] <rmatte> for instance, /usr/local/zenoss/zenoss/perf/Devices/lynchburg0111/os/interfaces/SPAN RP

[16-Sep-2011 15:54:34] <rmatte> it cuts that off at SPAN

[16-Sep-2011 15:54:35] <Hackman238> We patched our code to replace all space with _

[16-Sep-2011 15:54:43] <rmatte> ah

[16-Sep-2011 15:54:52] <rmatte> bleh

[16-Sep-2011 15:54:52] <Hackman238> Hum....

[16-Sep-2011 15:54:53] <rmatte> lol

[16-Sep-2011 15:54:58] <Hackman238> let me figure real quick

[16-Sep-2011 15:55:16] <Hackman238> Whats your current find commend

[16-Sep-2011 15:55:17] <Hackman238> ?

[16-Sep-2011 15:55:29] <rmatte> I think the issue may be in the script itself

[16-Sep-2011 15:55:32] <rmatte> you have...

[16-Sep-2011 15:55:34] <rmatte> `rrdtool dump $ARGV[0] > $tempfile`;

[16-Sep-2011 15:55:45] <rmatte> you might need quotes around the ARGV or something

[16-Sep-2011 15:55:54] <rmatte> my command is...

[16-Sep-2011 15:56:08] <rmatte> for i in `find $ZENHOME/perf/Devices/ -name ifOut*.rrd`; do /mspfiles/scripts/tools/rrd/removespikes/removespikes.pl "$i"; done

[16-Sep-2011 15:56:22] <rmatte> and I have a separate one for ifIn

[16-Sep-2011 15:59:05] <rmatte> it could also be because the script uses $ARGV[0] which probably only captures the first words before a space

[16-Sep-2011 15:59:14] <rmatte> anything after the space becomes $ARGV[1]

[16-Sep-2011 15:59:28] <Hackman238> find $ZENHOME/perf/Devices/ -name ifOut*.rrd -exec '/mspfiles/scripts/tools/rrd/removespikes/removespikes.pl' {} \;

[16-Sep-2011 15:59:45] <rmatte> k I'll give it a go, but it might not make a difference

[16-Sep-2011 16:00:26] <Hackman238> rmatte: If not, we'll tweak the script

[16-Sep-2011 16:00:28] <rmatte> now I get: ERROR: usage rrdtool dump [--no-header|-n] file.rrd [file.xml]

[16-Sep-2011 16:00:42] <rmatte> though it does work for some files

[16-Sep-2011 16:00:42] <Hackman238> rmatte: Gotcha- it didnt get a file name

[16-Sep-2011 16:01:12] <rmatte> No peaks found.!

[16-Sep-2011 16:01:12] <rmatte> No peaks found.!

[16-Sep-2011 16:01:13] <rmatte> ERROR: usage rrdtool dump [--no-header|-n] file.rrd [file.xml]

[16-Sep-2011 16:01:46] <Hackman238> find $ZENHOME/perf/Devices/ -name ifOut*.rrd | xargs -0 /mspfiles/scripts/tools/rrd/removespikes/removespikes.pl

[16-Sep-2011 16:02:07] <Hackman238> rmatte: Yeah if it says no peaks found, it worked but nothing was a peak

[16-Sep-2011 16:02:13] <rmatte> xargs: argument line too long

[16-Sep-2011 16:02:26] <Hackman238> rmatte: GRR!

[16-Sep-2011 16:02:27] <Hackman238> LOL

[16-Sep-2011 16:02:29] <rmatte> lol

[16-Sep-2011 16:02:42] <rmatte> I'm pretty sure it's your script causing the issue

[16-Sep-2011 16:02:47] <rmatte> since you're using $ARGV[0]

[16-Sep-2011 16:02:54] <rmatte> which cuts it after the first space it finds

[16-Sep-2011 16:03:06] <rmatte> I ran in to a similar issue with a bash script not too long ago

[16-Sep-2011 16:03:41] <rmatte> perhaps just using $ARGV would work

[16-Sep-2011 16:03:48] <Hackman238> rmatte: Try it

[16-Sep-2011 16:04:07] <Hackman238> rmatte: Or need to get quotes around the path results

[16-Sep-2011 16:04:39] <rmatte> right, but does perl even recognize quotes in arguments?

[16-Sep-2011 16:04:48] <rmatte> I already tried that by doing \"$i\"

[16-Sep-2011 16:04:52] <rmatte> and "$i"

[16-Sep-2011 16:04:56] <rmatte> neither worked

[16-Sep-2011 16:05:00] <rmatte> so it has to be the script

[16-Sep-2011 16:05:59] <Hackman238> Ah ha!

[16-Sep-2011 16:06:10] <rmatte> ?

[16-Sep-2011 16:06:22] <Hackman238> for i in `find $ZENHOME/perf/Devices/ -name ifOut*.rrd`; do  /mspfiles/scripts/tools/rrd/removespikes/removespikes.pl "'$i'"; done

[16-Sep-2011 16:06:34] <Hackman238> adds single quotes to path to make it one arg

[16-Sep-2011 16:06:36] <Hackman238> LOL

[16-Sep-2011 16:06:57] <rmatte> I'll give it a go

[16-Sep-2011 16:07:29] <rmatte> nope, same problem

[16-Sep-2011 16:07:45] <Hackman238> rmatte: Dam

[16-Sep-2011 16:07:47] <Hackman238> *damn

[16-Sep-2011 16:08:14] <Hackman238> This is what our paths look like: /opt/zenoss/perf/Devices/**DEVICE**/os/interfaces/GigabitEthernet1_2/ifOutErrors_ifOutErrors.rrd

[16-Sep-2011 16:08:20] <rmatte> is $ARGV[] valid?

[16-Sep-2011 16:08:39] <Hackman238> rmatte: Its probably an array

[16-Sep-2011 16:08:43] <rmatte> it's complaining about doing $ARGV.old for a filename

[16-Sep-2011 16:08:51] <rmatte> but it's ok with $ARGV[0].old

[16-Sep-2011 16:09:00] <Hackman238> rmatte: I'd definately change it back to $ARGV[0]

[16-Sep-2011 16:09:09] <rmatte> it's back to argv[0] now

[16-Sep-2011 16:09:25] <Hackman238> rmatte: Still doesnt work?

[16-Sep-2011 16:09:25] <rmatte> but it's not reading all of the values in to 0, there has to be a way to get the proper values out of it

[16-Sep-2011 16:09:30] <rmatte> but like I said, not a perl guy

[16-Sep-2011 16:09:36] <Hackman238> rmatte: Same ehre

[16-Sep-2011 16:09:42] <rmatte> ah

[16-Sep-2011 16:09:45] <Hackman238> rmatte: This is a question for pete

[16-Sep-2011 16:09:51] <Hackman238> rmatte: I'm the python guy

[16-Sep-2011 16:09:54] <Hackman238> LOL

[16-Sep-2011 16:09:55] <rmatte> lol

[16-Sep-2011 16:09:59] <rmatte> well is pete around?

[16-Sep-2011 16:10:10] <Hackman238> rmatte: 'fraid not. Bounced early.

[16-Sep-2011 16:10:18] <Hackman238> rmatte: Let me find Rich, he might know

[16-Sep-2011 16:10:19] <rmatte> bleh

[16-Sep-2011 16:10:23] <Hackman238> rmatte: PHP guy

[16-Sep-2011 16:10:36] <Hackman238> rmatte: (equally annoying lang)

[16-Sep-2011 16:10:45] <Hackman238> BRB

[16-Sep-2011 16:10:55] <jmp242> perl is line noise to me, but Devin likes it

[16-Sep-2011 16:11:00] <psyklone> rmatte - what's your perl question?

[16-Sep-2011 16:11:04] <hmp> perl ftw

[16-Sep-2011 16:11:37] <jmp242> everyone's left today here I suppose I ought to get going then...

[16-Sep-2011 16:12:11] <Hackman238> psyklone: for i in `find $ZENHOME/perf/Devices/ -name ifOut*.rrd`; do   /mspfiles/scripts/tools/rrd/removespikes/removespikes.pl "'$i'"; done

[16-Sep-2011 16:12:23] <Hackman238> ... to account for paths with spaces?

[16-Sep-2011 16:12:29] <Hackman238> psyklone: Also, did you get my PM?

[16-Sep-2011 16:13:01] <psyklone> Hackman238, i see it now - one sec

[16-Sep-2011 16:14:51] <Hackman238> psyklone: Thanks

[16-Sep-2011 16:16:38] <rmatte> I fixed it

[16-Sep-2011 16:16:46] <Hackman238> rmatte: What'd you do?

[16-Sep-2011 16:16:55] <rmatte> for i in `find $ZENHOME/perf/Devices/ -name ifOut*.rrd`; do  /mspfiles/scripts/tools/rrd/removespikes/removespikes.pl '$i'; done

[16-Sep-2011 16:16:59] <rmatte> single quotes

[16-Sep-2011 16:17:31] <Hackman238> rmatte: Ah-

[16-Sep-2011 16:18:23] <rmatte> actually that might not be working

[16-Sep-2011 16:18:28] <rmatte> it might just be passing a blank value to it

[16-Sep-2011 16:18:45] <rmatte> since if you do: echo '$i' you literally get $i

[16-Sep-2011 16:18:49] <rmatte> though maybe not

[16-Sep-2011 16:18:55] <rmatte> probably just echo functionality

[16-Sep-2011 16:19:11] <Hackman238> rmatte: Darn

[16-Sep-2011 16:19:24] <Hackman238> rmatte: I'm sure psyklone can figure a fix

[16-Sep-2011 16:19:36] <Hackman238> rmatte: If not, rich might have an answer

[16-Sep-2011 16:19:45] <rmatte> actually hmmm, I have an idea

[16-Sep-2011 16:21:08] <rmatte> I added a print line to the script and yeh, the single quotes don't appear to be working...

[16-Sep-2011 16:22:41] <Hackman238> rmatte: Alrighty.

[16-Sep-2011 16:26:16] <rmatte> this is ridiculous, no matter what I try it doesn't work

[16-Sep-2011 16:29:49] <rmatte> I swear, the perl parser for arguments is f**cked

[16-Sep-2011 16:30:05] <rmatte> according to the docs, passing quotes to it should make it read it in as a single argument

[16-Sep-2011 16:30:07] <rmatte> and it doesn't

[16-Sep-2011 16:30:57] <Hackman238> rmatte: Dave here suggested to pass the path through sed, replace all spaces with a # then replace all # in the script with space before usage

[16-Sep-2011 16:31:33] <Hackman238> rmatte: Sorry man- didnt mean to give you a first class ticket on rage train

[16-Sep-2011 16:31:39] <psyklone> how are you reading the arg in?

[16-Sep-2011 16:32:12] <rmatte> psyklone: I didn't actually write the script

[16-Sep-2011 16:32:38] <rmatte> Hackman238: do you mind if I fpaste it or is that a nono?

[16-Sep-2011 16:33:05] <Hackman238> rmatte: Go for it

[16-Sep-2011 16:33:18] <psyklone> i'm just wondering what the ARGV looks like where you're reading the input

[16-Sep-2011 16:33:26] <rmatte> psyklone: http://fpaste.org/OhVF/

[16-Sep-2011 16:33:29] <rmatte> there's the script

[16-Sep-2011 16:33:56] <rmatte> oh you know what

[16-Sep-2011 16:33:58] <rmatte> haha

[16-Sep-2011 16:34:01] <rmatte> I just realized something

[16-Sep-2011 16:34:18] <Hackman238> rmatte: ?

[16-Sep-2011 16:34:18] <rmatte> if I specify the quotes they do make it in to the argument, it must be the for loop that's messing it up at this point

[16-Sep-2011 16:34:23] <rmatte> let me try with -exec again

[16-Sep-2011 16:35:48] <rmatte> VICTORY

[16-Sep-2011 16:36:04] <rmatte> find $ZENHOME/perf/Devices/ -name ifOut*.rrd -exec /mspfiles/scripts/tools/rrd/removespikes/removespikes.pl \"{}\" \;

[16-Sep-2011 16:36:09] <rmatte> that works

[16-Sep-2011 16:36:13] <psyklone> sweet

[16-Sep-2011 16:36:31] <rmatte> now to run it through and see if it actually fixes my issue

[16-Sep-2011 16:36:33] <rmatte> thanks guys

[16-Sep-2011 16:37:53] <Hackman238> rmatte: Alrighty

[16-Sep-2011 16:37:55] <Hackman238> brb

[16-Sep-2011 16:38:13] <Hackman238> rmatte: Nice!

[16-Sep-2011 16:38:19] <Hackman238> rmatte: Glad its working

[16-Sep-2011 16:38:29] <rmatte> hmmmm

[16-Sep-2011 16:38:42] <Hackman238> rmatte: Oh no....too soon?

[16-Sep-2011 16:39:07] <rmatte> ERROR: creating '/usr/local/zenoss/zenoss/perf/Devices/wcFIhelsinki001/os/interfaces/10_100 Copper - Level 0x50d0001/ifOutUcastPkts_ifOutUcastPkts.rrd': File exists

[16-Sep-2011 16:39:21] <rmatte> /mspfiles/scripts/tools/rrd/removespikes/removespikes.pl: Unable to execute the rrdtool restore on "/usr/local/zenoss/zenoss/perf/Devices/wcFIhelsinki001/os/interfaces/10_100 Copper - Level 0x50d0001/ifOutUcastPkts_ifOutUcastPkts.rrd" - No such file or directory

[16-Sep-2011 16:39:33] <rmatte> strange

[16-Sep-2011 16:39:58] <Hackman238> rmatte: Wow...strange names

[16-Sep-2011 16:40:24] <rmatte> yeh, we have quite the assortment of interfaces

[16-Sep-2011 16:40:55] <rmatte> I just don't get why I got that error

[16-Sep-2011 16:41:01] <rmatte> it clearly had the full path right

[16-Sep-2011 16:41:41] <Hackman238> rmatte: need to escape the spaces?

[16-Sep-2011 16:42:03] <rmatte> nah, if I did it wouldn't have shown the full path in the error

[16-Sep-2011 16:42:22] <Hackman238> rmatte: true true

[16-Sep-2011 16:43:02] <rmatte> hmmm, it says file exists, maybe there's a deletion operation that's choking somewhere

[16-Sep-2011 16:44:18] <rmatte> I think this is failing...

[16-Sep-2011 16:44:20] <rmatte> rename($ARGV[0],"$ARGV[0].old");

[16-Sep-2011 16:46:54] <Hackman238> Hum

[16-Sep-2011 16:54:15] <rmatte> yeh that's exactly what's happening the rename is failing

[16-Sep-2011 16:55:05] <rmatte> so it's unable to do the restore and unable to do the cleanup

[16-Sep-2011 16:59:46] <Hackman238> rmatte: Any luck with a fix?

[16-Sep-2011 17:00:06] <rmatte> working on it

[16-Sep-2011 17:03:03] <rmatte> I tried doing it with the mv command instead via system()

[16-Sep-2011 17:03:05] <rmatte> mv: cannot stat `"/usr/local/zenoss/zenoss/perf/Devices/wcFIhelsinki001/os/interfaces/10_100 Copper - Level 0x50d0001/ifOutUcastPkts_ifOutUcastPkts.rrd"': No such file or directory

[16-Sep-2011 17:03:14] <rmatte> the stupid thing is, I just checked and that file does exist

[16-Sep-2011 17:03:19] <rmatte> so something is causing it not to

[16-Sep-2011 17:16:29] <Hackman238> rmatte: I'm sorry man

[16-Sep-2011 17:17:35] <rmatte> no big deal, I'll sort it out eventually

[16-Sep-2011 17:19:10] <Hackman238> Time to run

[16-Sep-2011 17:19:13] <Hackman238> Later man

[16-Sep-2011 17:21:07] <rmatte> later

[16-Sep-2011 17:44:54] kocolosk is now known as kocolosk|away

[16-Sep-2011 18:12:03] <rmatte> there, finally fixed the script

[16-Sep-2011 18:12:23] <rmatte> now to run it and see if it actually helps

[16-Sep-2011 18:26:48] <rmatte> now that I finally have this script working it's massively pimp

[16-Sep-2011 18:27:01] <rmatte> I might just cron it to run weekly or something

[16-Sep-2011 18:27:13] <rmatte> though that's probably a bad idea

[16-Sep-2011 21:41:32] <rmatte> well, the petabyte value is gone but now I have a terabyte negative value in there somewhere, time to break out my script for removing negative values from RRDs

[16-Sep-2011 21:41:58] <rmatte> I'm going to package this stuff up on Monday and write a wiki entry about it

[16-Sep-2011 21:42:30] <rmatte> between my script and the script that Shane sent me it's a pretty solid way to normalize values in interface RRDs

[16-Sep-2011 21:49:38] <rmatte> Normalizing: -1.6417750842e+15

[16-Sep-2011 21:49:42] <rmatte> I believe that would be it

[16-Sep-2011 21:57:50] kocoloskz is now known as kocolosk

[16-Sep-2011 21:59:48] <rmatte> yup, epic win

[16-Sep-2011 22:21:31] <bender183> lol

[17-Sep-2011 00:00:01] [disconnected at Sat Sep 17 00:00:01 2011]

[17-Sep-2011 00:00:02] [connected at Sat Sep 17 00:00:02 2011]

[17-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[17-Sep-2011 12:00:01] [disconnected at Sat Sep 17 12:00:01 2011]

[17-Sep-2011 12:00:02] [connected at Sat Sep 17 12:00:02 2011]

[17-Sep-2011 12:00:15] [zenoss-logger (logger bot) has joined #zenoss]

[17-Sep-2011 23:49:28] <Tux_Tux> Good evening, I have a question for the moment opennms to monitor use of network resources, and I have integrated with otrs system for tikets, I want to start using ZenoOS but find that tikets system and I can integrate online help, someone know of any thanks, forgive my English

[18-Sep-2011 00:00:01] [disconnected at Sun Sep 18 00:00:01 2011]

[18-Sep-2011 00:00:02] [connected at Sun Sep 18 00:00:02 2011]

[18-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[18-Sep-2011 00:12:30] <Tux_Tux> hi,, i have one question

[18-Sep-2011 08:31:07] kocolosk is now known as kocolosk|away

[18-Sep-2011 08:31:16] kocolosk|away is now known as kocolosk

[18-Sep-2011 08:58:32] kocolosk is now known as kocolosk|away

[18-Sep-2011 12:00:01] [disconnected at Sun Sep 18 12:00:01 2011]

[18-Sep-2011 12:00:01] [connected at Sun Sep 18 12:00:01 2011]

[18-Sep-2011 12:00:14] [zenoss-logger (logger bot) has joined #zenoss]

[18-Sep-2011 13:24:40] <Tux_Tux> hola

[18-Sep-2011 13:25:31] <Tux_Tux> hi, someone who can help me with a question about ZenoOS?

[18-Sep-2011 14:05:10] <Tux_Tux> pregunta, con que sistema de tickets zenoss se integra mejor y sea open source?

[18-Sep-2011 14:05:48] <Tux_Tux> question that zenoss ticketing system is better integrated and be open source?

[18-Sep-2011 15:19:09] tvincent_ is now known as tvincent

[18-Sep-2011 15:21:25] tvincent_ is now known as tvincent

[18-Sep-2011 22:51:17] <hikenboot1> hello how do i tell which port is recieving traps on in zenoss?

[18-Sep-2011 22:51:54] <hikenboot1> and secondly do i have to specify it on the remote machines sending to the traps

[18-Sep-2011 22:52:35] <hikenboot1> bbl to see replies going out for dinner...thanks in advance

[19-Sep-2011 00:00:01] [disconnected at Mon Sep 19 00:00:01 2011]

[19-Sep-2011 00:00:01] [connected at Mon Sep 19 00:00:01 2011]

[19-Sep-2011 00:00:15] [zenoss-logger (logger bot) has joined #zenoss]

[19-Sep-2011 04:46:46] <tsener> hello

[19-Sep-2011 04:48:31] <enkrypt> hello

[19-Sep-2011 04:48:59] <tsener>

[19-Sep-2011 04:49:13] <tsener> anyome tried  the freebsd zenoss port ?

[19-Sep-2011 04:49:22] <enkrypt> didn't even know there is one

[19-Sep-2011 05:12:33] <tsener>

[19-Sep-2011 06:01:25] <rcngeoff> hi...  I can't find this in the docs... how do I manually kick off a new zenmodeler for my entire network?

[19-Sep-2011 06:18:51] froztbyte is now known as froztstrawberryf

[19-Sep-2011 06:18:59] froztstrawberryf is now known as froztbyte

[19-Sep-2011 06:24:30] <enkrypt> it seems my zenrender process doesn't auto start on the hub

[19-Sep-2011 06:24:40] <enkrypt> what is the cleanest way to fix that?

[19-Sep-2011 07:24:35] <MariaKeys> hello fellas. does zenoss 3.2 run on rhel6?

[19-Sep-2011 07:25:01] <enkrypt> MariaKeys: zenoss is currently not supported on rhel6 unfortunately

[19-Sep-2011 07:25:51] <MariaKeys> thanks enkrypt

[19-Sep-2011 07:26:23] <MariaKeys> i find this rather strange.

[19-Sep-2011 07:26:32] <MariaKeys> redhat means usually software 10 years old.

[19-Sep-2011 07:26:42] <MariaKeys> and it has been out for 1 year.

[19-Sep-2011 07:28:22] <enkrypt> yeah it's a bit unfortunate

[19-Sep-2011 07:28:34] <enkrypt> even zenoss 4.x (the "new generation") is meant for RHEL5

[19-Sep-2011 07:29:17] <MariaKeys> hmm. thanks again.

[19-Sep-2011 07:29:26] <enkrypt> nyeates / rmatte: I have a feeling OSProcess monitoring is broken for linux systems on zenoss 3.2.0, can anyone confirm that? in 3.1.0 it worked fine, in 3.2.0 it breaks down

[19-Sep-2011 07:41:55] <Jane_Curry> hikenboot1: The zentrap daemon receives traps on UDP/162

[19-Sep-2011 07:42:40] <Jane_Curry> hikenboot1: Any agents sending traps to Zenoss will typically use UDP/162 but the agent must be configured with the zenoss system as (one of) their trap destinations

[19-Sep-2011 08:29:07] <jmp242> rcngeoff: you can use 'zenmodeler run'

[19-Sep-2011 08:29:24] <jmp242> Anyone seen this thread and has a comment? message/61414#61414

[19-Sep-2011 08:38:58] <rcngeoff> @jmp242: thanks

[19-Sep-2011 08:42:25] <dec3pti01> so not that I can create copies of templates ... how come sometimes the copies do not appear with the green icon to the left of the class path ?

[19-Sep-2011 08:42:49] <dec3pti01> do I need to do something to make it green .. I believe in the past I just waited

[19-Sep-2011 08:43:58] <dec3pti01> he nevermind

[19-Sep-2011 09:33:01] kocolosk|away is now known as kocolosk

[19-Sep-2011 09:40:19] <hikenboot1> can someone check out my nexentastor snmpd.conf and tell  me what they think seems wrong with it http://pastebin.ca/2081278

[19-Sep-2011 09:43:15] <SDuensin> Morning.

[19-Sep-2011 09:51:50] <Jane_Curry> hikenboot1:  So do you have oid configured in zSnmpCommunity for this device in Zenoss?

[19-Sep-2011 09:53:17] <Jane_Curry> hikenboot: Also, your snmpd.conf has no trapsink statement to send traps to your Zenoss

[19-Sep-2011 09:53:33] <hikenboot1> Jane_Curry: Thanks in advance no i have nothing config in zenoss I just had it detect devices and it shows the ip of the nexentasor

[19-Sep-2011 09:54:19] <Jane_Curry> try setting the zSNMPCommunity to oid and remodel the device - you should get lots more stuff

[19-Sep-2011 09:58:12] <Jane_Curry> With "new" pages, browser access is controlled by statements in zcml files

[19-Sep-2011 09:58:54] <Jane_Curry> If I want to override "permission=" statements defined in Core, by a ZenPack, can I do that? If so, how?

[19-Sep-2011 10:01:09] <rmatte> enkrypt: all that pack is is a basic template with two command datasources.  If it's exhausting your zenhub then there's either a bug or you don't have sufficient hardware on that server.

[19-Sep-2011 10:13:10] jb____ is now known as jb

[19-Sep-2011 10:13:55] <hikenboot1> thanks Jane_Curry: I will look into what you say

[19-Sep-2011 10:16:37] <hikenboot1> Jane_Curry: I am new to snmp but it appears I need some kind of mib walker to determine what MIB I need to use...is this correct

[19-Sep-2011 10:16:49] <hikenboot1> in this case its nexentastor based on solaris

[19-Sep-2011 10:17:01] <hikenboot1> open solaris

[19-Sep-2011 10:24:54] <jb> you need to walk the MIB to determine what OID's you would like to poll, yes.

[19-Sep-2011 10:43:42] <Jane_Curry> hikenboot1: unless you want to do fancy stuff, you shouldn't need a MIB browser.  There are lots of standard queries built-in to both Zenoss and your agent

[19-Sep-2011 10:44:19] <Jane_Curry> that should give you system stuff, networks, filesystems, processes.  Looks like your agent is based on net-snmp which is VERY common

[19-Sep-2011 10:46:26] <hikenboot1> thanks Jane_Curry

[19-Sep-2011 10:59:37] <dpetzel> any chance someone has a sec to look at message/61402#61402 ?

[19-Sep-2011 11:01:04] <rmatte> dpetzel: yeh, so the issue is how to get data from a datasource that contains a space?

[19-Sep-2011 11:01:14] <dpetzel> rmatte: yeah

[19-Sep-2011 11:01:26] <dpetzel> I got it to work with some hackery of core files, but probably not the best approach

[19-Sep-2011 11:03:02] <rmatte> well, since you're only getting one value you don't need the square brackets around it for starters

[19-Sep-2011 11:03:10] <rmatte>     print d.getRRDValues(['ServiceName HTTP totalServicedSum'],

[19-Sep-2011 11:03:12] <rmatte>                              start=startDate, end=endDate, function=how)

[19-Sep-2011 11:03:27] <rmatte>     print d.getRRDValues('ServiceName HTTP totalServicedSum',start=startDate, end=endDate, function=how)

[19-Sep-2011 11:04:17] <rmatte> and I could swear that I've managed to get data perfectly fine from datasources that have spaces in them in the past

[19-Sep-2011 11:04:20] <rmatte> I'll find out and show you

[19-Sep-2011 11:04:25] <rmatte> find one*

[19-Sep-2011 11:04:41] <dpetzel> rmatte: OK cool thanks, yeah I can't fathom why it would be aproble. Graphs render just fine, reports work

[19-Sep-2011 11:04:49] <dpetzel> so I think I'm just missing something obvious

[19-Sep-2011 11:05:08] fOrsberg is now known as forsberg

[19-Sep-2011 11:05:18] <dpetzel> in case it matters this is on 3.1

[19-Sep-2011 11:05:27] <rmatte> It shouldn't

[19-Sep-2011 11:05:40] <rmatte> if I can get it working in 2.5 then it'd be a bug in 3.x if it doesn't work anymore

[19-Sep-2011 11:08:02] <rmatte> waiting for zendmd to load...

[19-Sep-2011 11:08:35] <dpetzel> rmatte: without brackets, same result

[19-Sep-2011 11:10:54] <rmatte> hmmm, you're right, it does throw an error... but Zenoss obviously has some way around it, just need to figure out what it does

[19-Sep-2011 11:11:13] <dpetzel> rmatte: agreed, that way around is how I got stumped

[19-Sep-2011 11:11:58] <jmp242> Speaking of forum posts - anyone have a chance to look at: message/61414#61414

[19-Sep-2011 11:12:02] <enkrypt> nyeates and rmatte I've opened a support ticket for the OSProcess monitoring issue: http://dev.zenoss.com/trac/ticket/7870

[19-Sep-2011 11:13:06] forsberg is now known as fOrsberg

[19-Sep-2011 11:15:37] <rmatte> hmmmm, I think it's throwing the error for me because I'm trying it against a datasource rather than a data point

[19-Sep-2011 11:15:46] <rmatte> so my results aren't conclusive

[19-Sep-2011 11:16:05] <rocket> jmp242: you can ignore the warning for 61414

[19-Sep-2011 11:16:06] <dpetzel> rmatte: doesn getRRDValues require the passing of a data source, not data point?

[19-Sep-2011 11:16:23] <rocket> jmp242: basically its a warning saying its trying to remove something thats already been removed

[19-Sep-2011 11:16:32] <rmatte> dpetzel: only if there's only one datapoint under the datasource

[19-Sep-2011 11:16:44] <rmatte> dpetzel: you can have datasources with several datapoints under them

[19-Sep-2011 11:16:48] <rmatte> such as command datasources

[19-Sep-2011 11:16:57] <rmatte> at which point using the datasource name can not and will not work

[19-Sep-2011 11:16:59] <dpetzel> rmatte: nod, yeah I have a couple liek taht too (but have not tried to report on them yet)

[19-Sep-2011 11:17:26] <rmatte> by the way, if you're the one who put the space in the datasource name in the first place, bad practice

[19-Sep-2011 11:17:33] <rmatte> you should avoid spaces at all costs

[19-Sep-2011 11:17:51] <dpetzel> rmatte: yes, that lessoned has been learned a few times over but the damage has been done

[19-Sep-2011 11:17:55] <rmatte> you should use camelcase, like... PoisonQueueLength

[19-Sep-2011 11:18:04] <rmatte> k

[19-Sep-2011 11:18:33] <rmatte> have you tried simply substituting spaces with underscores?

[19-Sep-2011 11:18:52] <rmatte> print d.getRRDValues('ServiceName_HTTP_totalServicedSum',start=startDate, end=endDate, function=how)

[19-Sep-2011 11:19:04] <dpetzel> yeah I tried that, it never finds the data source then

[19-Sep-2011 11:19:26] <dpetzel> since getRRDValues is enumerating based on dsname

[19-Sep-2011 11:19:38] <rmatte> ServiceName HTTP totalServicedSum <--- Is that the entire name of the datasource

[19-Sep-2011 11:19:41] <dpetzel> I also tried escaping the space with no love

[19-Sep-2011 11:19:54] <dpetzel> yeah that is the full name of the data source

[19-Sep-2011 11:19:56] <rmatte> or is totalServicedSum the datapoint name and ServiceName HTTP is the datasource name?

[19-Sep-2011 11:20:41] <rmatte> these are important details

[19-Sep-2011 11:20:56] <dpetzel> Service HTTP totalServicedSum = Data source name

[19-Sep-2011 11:20:57] <rmatte> is it an SNMP type datasource or a command datasource?

[19-Sep-2011 11:21:03] <dpetzel> totalServicedSum = data point name

[19-Sep-2011 11:21:10] <dpetzel> Service HTTP totalServicedSum_totalServicedSum.rrd == RRD file

[19-Sep-2011 11:21:16] <dpetzel> this is aJMX data source

[19-Sep-2011 11:21:22] <rmatte> ok, simple then, you're over-complicating it... do this...

[19-Sep-2011 11:21:34] <rmatte> print d.getRRDValues('totalServicedSum',start=startDate, end=endDate, function=how)

[19-Sep-2011 11:22:02] <dpetzel> here is the catch... with that though I think I have multiple datasources on the same box with that datapoint name

[19-Sep-2011 11:22:07] <rmatte> provided there aren't more than one datapoint with that exact name

[19-Sep-2011 11:22:15] <dpetzel> Service HTTP totalServicedSum_totalServicedSum.rrd

[19-Sep-2011 11:22:15] <rmatte> ah

[19-Sep-2011 11:22:15] <dpetzel> Service1 HTTP totalServicedSum_totalServicedSum.rrd

[19-Sep-2011 11:22:15] <dpetzel> Service2 HTTP totalServicedSum_totalServicedSum.rrd

[19-Sep-2011 11:22:23] <rmatte> I see

[19-Sep-2011 11:22:29] <dpetzel> we stack 5-10 java apps on the same box

[19-Sep-2011 11:25:16] <dpetzel> I think I might be able to extend RRDView and override getRRDValues to bend to my wishes, but that feels like Im overcomplicating for sure

[19-Sep-2011 11:25:39] <rmatte> you may have to actually iterate through each datapoint in dmd and do the getRRDValue directly against the datapoints instead of against the device

[19-Sep-2011 11:25:41] <jmp242> thanks rocket

[19-Sep-2011 11:25:50] <rmatte> that's probably how they do it in Zenoss

[19-Sep-2011 11:27:15] <dpetzel> rmatte: OK, Ill try that angle see how I make out

[19-Sep-2011 11:27:19] <rocket> we iterate through the datapoints on the device if I recall correctly

[19-Sep-2011 11:27:32] <rmatte> yeh

[19-Sep-2011 11:28:21] <rocket> there might be a catalog that returns just the datapoints with a specific name

[19-Sep-2011 11:28:38] <rocket> I believe we use the catalogs for optimization instead of a for loop

[19-Sep-2011 11:29:02] <rmatte> I think there's some function that does it, just trying to remember which

[19-Sep-2011 11:30:26] <rmatte> d.getRRDDataPoints to get the list

[19-Sep-2011 11:30:42] <rmatte> not sure about what would return a catalog

[19-Sep-2011 11:46:29] <jbk> is there a way to create a per-instance threshold (for example, an effectively inactive filesystem might be at 90% full, but don't want it to trigger an event)?

[19-Sep-2011 11:48:36] <rmatte> jbk: yes, you create a local copy of the template then change the threshold value on it

[19-Sep-2011 11:49:32] <jbk> doesn't that only work at the device level?

[19-Sep-2011 11:49:51] <jbk> at not the component?

[19-Sep-2011 11:50:05] <rmatte> no, it works at the component level

[19-Sep-2011 11:50:12] <rmatte> there are device templates and component templates

[19-Sep-2011 11:50:37] <rmatte> they all look the same, the difference is the Python Class property in the template

[19-Sep-2011 11:51:38] <rmatte> it'll be set to something like Products.ZenModel.FileSystem

[19-Sep-2011 11:51:44] <rmatte> depending on the class the component belongs to

[19-Sep-2011 11:51:54] <rmatte> and it'll automatically apply the template to the component with the matching class

[19-Sep-2011 12:00:01] [disconnected at Mon Sep 19 12:00:01 2011]

[19-Sep-2011 12:00:02] [connected at Mon Sep 19 12:00:02 2011]

[19-Sep-2011 12:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[19-Sep-2011 12:27:42] <dhopp> ok..can somebody do a sanity check on my RRA definitions?  The RRD files are smaller than I was expecting..

[19-Sep-2011 12:27:51] <dhopp> Assuming a 5 minute interval...

[19-Sep-2011 12:28:01] <dhopp> RRA:AVERAGE:0.5:1:8640

[19-Sep-2011 12:28:01] <dhopp> RRA:AVERAGE:0.5:3:35040

[19-Sep-2011 12:28:01] <dhopp> RRA:AVERAGE:0.5:6:36000

[19-Sep-2011 12:28:01] <dhopp> RRA:MAX:0.5:3:35040

[19-Sep-2011 12:28:01] <dhopp> RRA:MAX:0.5:6:36000

[19-Sep-2011 12:28:15] <dhopp> that should do 5 minutes for a month, 15 minutes for 1 year and 30 minutes for 2 years..right?

[19-Sep-2011 12:28:25] <dhopp> well technically a little more then 2 years on the 30 minutes

[19-Sep-2011 12:30:56] <rmatte> one sec, checking...

[19-Sep-2011 12:33:31] <rmatte> how exactly did you do the match for that first line?

[19-Sep-2011 12:33:35] <rmatte> math*

[19-Sep-2011 12:34:53] <rmatte> as far as I'm aware what you want to do is the following...

[19-Sep-2011 12:35:57] <rmatte> first you find out how many cycles that is per hour: 60 / 5 = 12 x 24 = 288 x 30 = 8640

[19-Sep-2011 12:36:01] <rmatte> so yeh, that one's good

[19-Sep-2011 12:36:19] <rmatte> second one is 15, so...

[19-Sep-2011 12:37:13] <rmatte> 60 / 15 = 4 x 24 = 96 x 365 = 35040

[19-Sep-2011 12:37:32] <rmatte> but it'll consolidate the data, so it'll no longer be high resolution

[19-Sep-2011 12:38:26] <rmatte> the last one will be: 60 / 30 = 2 x 24 = 48 x 365 = 17520 x 2 = 35040

[19-Sep-2011 12:38:38] <rmatte> because of the massive consolidation of data though, the file will be quite small

[19-Sep-2011 12:39:48] <rmatte> if you wanted high resolution for an entire year you'd do something like: RRA:AVERAGE:0.1:1:525600

[19-Sep-2011 12:39:58] <rmatte> or high resolution for 2 years: RRA:AVERAGE:0.1:1:1051200

[19-Sep-2011 12:41:33] <rmatte> but the RRDs will be hundreds of megs

[19-Sep-2011 12:41:44] <rmatte> well maybe not

[19-Sep-2011 12:41:50] <rmatte> more like 16MB each or something

[19-Sep-2011 12:42:06] <rmatte> (I did that recently on one server)

[19-Sep-2011 12:43:40] <rmatte> yeh, just checked, for RRA:AVERAGE:0.1:1:1051200 which is 2 years worth of high resolution data at 1 minute cycles the RRDs come out to 17MB each

[19-Sep-2011 12:43:52] <rmatte> so don't expect your RRDs to be huge

[19-Sep-2011 12:58:53] <dhopp> yeah I knew it was going to be small..but I was expecting 3-4 MB..but it's only 1.2 MB

[19-Sep-2011 12:59:13] <dhopp> I don't want 2 years of 5 minute..but I want 2 years of 30 minutes

[19-Sep-2011 13:00:40] <nyeates> You guys see that there is a beta out?

[19-Sep-2011 13:01:04] <nyeates> please do go test it and let us know anything you find

[19-Sep-2011 13:01:54] <nyeates> i gotta leave though, later

[19-Sep-2011 13:07:07] <hikenboot1> where do i find a list of community strings for solaris snmp?

[19-Sep-2011 13:07:15] <rmatte> dhopp: 2 years of 30 minutes is going to by tiny

[19-Sep-2011 13:07:15] <hikenboot1> and for windows? and Linux?

[19-Sep-2011 13:07:17] <rmatte> so that's expected

[19-Sep-2011 13:07:31] <rmatte> hikenboot1: there is no list, you set the string yourself

[19-Sep-2011 13:07:37] <rmatte> the default is always "public"

[19-Sep-2011 13:07:54] <rmatte> the string is like a password for your snmp agents

[19-Sep-2011 13:08:07] <hikenboot1> oh so the community string isnt some  nubmer like 1.3.5.7.8.3.23.2?

[19-Sep-2011 13:08:14] <rmatte> no, that's an SNMP OID

[19-Sep-2011 13:08:27] <hikenboot1> ok where do i find a list of OIDs?

[19-Sep-2011 13:08:43] <rmatte> hikenboot1: Zenoss already has pre-built templates for Linux and Solaris

[19-Sep-2011 13:08:59] <rmatte> which include the most common OIDs

[19-Sep-2011 13:09:03] <hikenboot1> got a doc on using the prebuilt templates?

[19-Sep-2011 13:09:22] <rmatte> Read the Zenoss Admin Guide top to bottom so that you understand how to use Zenoss in general

[19-Sep-2011 13:09:32] <rmatte> it includes detailed info about templates

[19-Sep-2011 13:09:39] <hikenboot1> thanks

[19-Sep-2011 13:09:43] <rmatte> http://downloads.sourceforge.net/zenoss/Zenoss_Administration_06-102010-3.0-v03.pdf

[19-Sep-2011 13:10:50] <rmatte> also, I assume you're a new Zenoss user, did you just install version 3.2?

[19-Sep-2011 13:10:58] <rmatte> if so, it has bugs in it that need to be patched

[19-Sep-2011 13:11:10] <rmatte> you should ssh to the server, become the zenoss user, then issue the following commands...

[19-Sep-2011 13:11:11] <dhopp> rmatte:  3.2.1 beta apparently is out now

[19-Sep-2011 13:11:21] <rmatte> dhopp: no it's not, the beta is out

[19-Sep-2011 13:11:28] <rmatte> zenpatch 26334

[19-Sep-2011 13:11:31] <rmatte> zenpatch 26335

[19-Sep-2011 13:11:34] <rmatte> zenpatch 2633

[19-Sep-2011 13:11:38] <rmatte> erm

[19-Sep-2011 13:11:38] <dhopp> rmatte: hence why I said "3.2.1 beta"

[19-Sep-2011 13:11:45] <rmatte> zenpatch 26337

[19-Sep-2011 13:12:08] <rmatte> ah, I'm just sight reading due to multitasking, missed the "beta"

[19-Sep-2011 13:12:29] <rmatte> anyways, I'm not going to recommend a beta to a new user even if it is just 3.2 with some patches applied

[19-Sep-2011 13:13:49] <rmatte> 3.2 shouldn't even be released at this point

[19-Sep-2011 13:13:54] <rmatte> that was a huge screwup on their part

[19-Sep-2011 13:57:02] <dhopp> is there a big draw back to having big RRD files (say bigger than 5 MB or so)?

[19-Sep-2011 13:57:18] <dhopp> does that add time to the rrdupdate cycle?

[19-Sep-2011 14:00:32] <dpetzel> rmatte: Thanks for your help earlier, got it working: message/61402

[19-Sep-2011 14:02:36] fOrsberg is now known as forsberg

[19-Sep-2011 14:11:01] <rmatte> dhopp: it adds a bit of time when reading from them I believe, but probably not when writing to them

[19-Sep-2011 14:11:14] <rmatte> dpetzel: cool, no problem

[19-Sep-2011 14:19:34] <dhopp> rmatte:  true..you would need a ton of RAM to allow the OS to cache them for reading...

[19-Sep-2011 14:19:34] <dhopp> hrm

[19-Sep-2011 14:27:52] <jb> i need valid address information for each location for the maps portlet to work correctly, right?

[19-Sep-2011 14:28:12] <jb> so, if I go to the location, and click edit, and populate the address.. everything should work correctly?

[19-Sep-2011 15:35:16] <rmatte> jb: yes, though sometimes it'll hiccup on certain addresses

[19-Sep-2011 15:35:22] <rmatte> you have to play around with it until it loads correctly

[19-Sep-2011 15:35:40] <rmatte> generally it should be fairly problem free though

[19-Sep-2011 15:35:56] <rmatte> hmmm, you actually just reminded me...

[19-Sep-2011 15:36:04] <rmatte> the new google maps API no longer requires keys

[19-Sep-2011 15:36:16] <rmatte> I wonder when Zenoss will update to using it and get rid of this API key nonsense

[19-Sep-2011 15:36:21] <rmatte> it's been out for a while

[19-Sep-2011 15:52:28] <jbk> hmm.. i suspect this should be returning a long, not int

[19-Sep-2011 15:52:37] <jbk> (bundled FileSystem.py)

[19-Sep-2011 15:54:15] <jbk> though the vm image is a couple months old.. might have been fixed since then

[19-Sep-2011 15:54:15] <jb> rmatte: can we use  GPS coordinates instead of a physical address?

[19-Sep-2011 15:54:33] <jbk> (i'm seeing negative sizes, even though all the snmp types are counter64)

[19-Sep-2011 15:56:34] <jb> looks like it.

[19-Sep-2011 15:56:45] <jb> yay.. i have devices in the middle of nowhere that I only can locate via GPS.

[19-Sep-2011 15:56:52] <jb> so this will be nice if it works

[19-Sep-2011 16:04:23] <jbk> looks like that fixed it

[19-Sep-2011 16:07:44] <rmatte> lol

[19-Sep-2011 16:08:11] <rmatte> monitoring buoys floating around the ocean are you?

[19-Sep-2011 16:08:53] <rmatte> that would actually be an interesting idea, get gps coordinates from devices that actually move and auto-update them on the google maps

[19-Sep-2011 16:12:30] <jb> nah, wireless tower sites.

[19-Sep-2011 16:12:37] <rmatte> ah

[19-Sep-2011 16:12:45] <dhopp> done

[19-Sep-2011 16:12:47] <dhopp> oops

[19-Sep-2011 16:13:03] <jb> however, I can't get it working.. specifying a coordinate like "N37 40.4230, W85 17.9750"

[19-Sep-2011 16:13:12] <jb> which is what I would type in Google Maps to display the location.

[19-Sep-2011 16:13:40] <rmatte> yeh, it's quite picky like I said

[19-Sep-2011 16:14:00] <rmatte> sometimes I have to make it just the city name for some places in europe or it won't work

[19-Sep-2011 16:14:02] <rmatte> it's weird

[19-Sep-2011 16:14:55] <jbk> is there a public browsable scm repo -- i'd like to see if this fix I did has been fixed in newer versions

[19-Sep-2011 16:16:05] <jb> rmatte: i wonder if I am specifyin them correctly.

[19-Sep-2011 16:16:24] <jb> this thread indicates they are specified differently.. thread/8275

[19-Sep-2011 16:20:51] <rmatte> well try it that way

[19-Sep-2011 16:21:32] <rmatte> "Put a comma between lat and long:

[19-Sep-2011 16:21:33] <rmatte> 39.751036, -94.840482"

[19-Sep-2011 16:21:34] <jb> well, I'm trying to figure out how to convert them

[19-Sep-2011 16:21:44] <rmatte> ah

[19-Sep-2011 16:22:33] <jb> hr,.. http://transition.fcc.gov/mb/audio/bickel/DDDMMSS-decimal.html

[19-Sep-2011 16:23:21] <jb> but, im not sure what format I have is.

[19-Sep-2011 16:39:26] <jb> yeah, still doesn't work in that format

[19-Sep-2011 16:39:28] <jb> bleh

[19-Sep-2011 16:40:45] <jb> rmatte: would you mind trying a GPS coordinate on a location on your install? 37.673706, -85.299581

[19-Sep-2011 16:53:04] <dhopp> I know

[19-Sep-2011 16:53:06] <dhopp> err

[19-Sep-2011 16:53:12] <dhopp> damn it..too many windows open

[19-Sep-2011 16:59:25] kocolosk is now known as kocolosk|away

[19-Sep-2011 17:44:29] <illsci> hey I'm getting this warning for my hosts... I'm not 100% sure why... but I think I might know why but I'm not sure how to fix it...

[19-Sep-2011 17:44:55] <illsci> Unable to read processes on device 10.blah; error: first argument must be a string or compiled pattern

[19-Sep-2011 17:45:18] <illsci> so it's using the ip instead of some string....?

[19-Sep-2011 17:45:34] <illsci> The Device's name is an ip at the moment

[19-Sep-2011 17:46:04] <illsci> Is this just a matter of me changing the devices name from an ip to a string

[19-Sep-2011 18:15:06] <crazed_> what's zenpatch 26337 do

[19-Sep-2011 18:15:12] <crazed_> i just upgraded and applied those first two

[19-Sep-2011 18:15:18] <crazed_> to fix graphing

[19-Sep-2011 18:15:29] crazed_ is now known as crazed

[19-Sep-2011 18:26:04] <crazed> hm i have an interesting issue

[19-Sep-2011 18:26:39] <crazed> where modeler picks up processes, but seems i get errorsregarding processes being down

[19-Sep-2011 18:26:43] <crazed> even though they are definitely up

[19-Sep-2011 18:26:55] <crazed> anyone see that on 3.2? i've been searching around

[19-Sep-2011 19:25:16] forsberg is now known as fOrsberg

[19-Sep-2011 19:46:32] kocolosk|away is now known as kocolosk

[19-Sep-2011 20:54:30] <illsci> y0 y0 y0

[19-Sep-2011 20:54:32] <illsci> what's up

[19-Sep-2011 21:03:42] <crazed> not much

[19-Sep-2011 21:03:48] <crazed> having issues with zenoss process matching

[19-Sep-2011 21:06:55] <kocolosk> crazed: i've seen zenprocess go a bit crazy when the command-line args used to launch the process change

[19-Sep-2011 21:07:35] <kocolosk> it goes into a bit of a death spiral setting and clearing events for the process.  Restarting the daemon fixes it

[19-Sep-2011 21:07:54] <crazed> eh this is for 3.2

[19-Sep-2011 21:07:57] <crazed> it started after i upgraded

[19-Sep-2011 21:08:01] <crazed> from 3.1

[19-Sep-2011 21:09:24] <crazed> it seems to mostly be processes with [ ] (brackets) in the name

[19-Sep-2011 21:09:43] <illsci> have you ever seen this? Unable to read processes on device 10.113.42.141; error: first argument must be string or compiled pattern

[19-Sep-2011 21:09:53] <illsci> is that fixed by a zenpatch of some sort?

[19-Sep-2011 21:10:06] <illsci> my device's name is an ip addres...

[19-Sep-2011 21:10:22] <illsci> I don't see what the issue is... it looks like it doesn't like the type of one of the properties...

[19-Sep-2011 22:34:03] <tvincent> anyone know how to get the queue length via DMD?

[19-Sep-2011 23:12:04] <JackSafro> any devs live here?  seems like lots of people are having problems with the regex process matching on the 3.2.0 release.  is there a zenpatch that i am missing in my google searches to fix this?

[19-Sep-2011 23:12:54] <JackSafro> the issue seems to stem from Products/ZenRRD/zenprocess.py and the match() function

[19-Sep-2011 23:20:09] <crazed> JackSafro: i'm having the same problem

[19-Sep-2011 23:20:59] <crazed>             if self.digest and digest != self.digest:

[19-Sep-2011 23:21:04] <crazed> i believe that line is the problem

[19-Sep-2011 23:21:19] <crazed> well, or the stuff in self.digest is sometimes wrong

[19-Sep-2011 23:23:17] <crazed> example, i have a process 'mogilefsd [delete]', and digest becomes some md5 value built off of 'args', while self.digest is set to '_delete'

[19-Sep-2011 23:24:18] <crazed> so the match is always set to false :/

[19-Sep-2011 23:25:20] <crazed> i'm not sure where a process' digest gets set originally

[19-Sep-2011 23:30:15] <crazed> it appears that when args is '', the digest gets set to d41d8cd98f00b204e9800998ecf8427e

[19-Sep-2011 23:42:39] <JackSafro> d41d8cd98f00b204e9800998ecf8427e is the md5sum of null

[19-Sep-2011 23:53:10] <crazed> yes

[20-Sep-2011 00:00:01] [disconnected at Tue Sep 20 00:00:01 2011]

[20-Sep-2011 00:00:02] [connected at Tue Sep 20 00:00:02 2011]

[20-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[20-Sep-2011 00:58:47] kocolosk is now known as kocolosk|away

[20-Sep-2011 03:41:10] fOrsberg is now known as forsberg

[20-Sep-2011 03:45:10] <hmp> i have a problem with device being snmp-ed and transform error being shown

[20-Sep-2011 03:45:41] <hmp> the device is no longer modeled in zenoss, as i have changed the ip on that device

[20-Sep-2011 03:45:54] <hmp> pardon, it is modeled, but not with that ip

[20-Sep-2011 03:46:36] <hmp> would reindexing the zodb solve the problem?

[20-Sep-2011 07:17:05] forsberg is now known as fOrsberg

[20-Sep-2011 07:22:34] <Jane_Curry> crazed: 26337 fixes zentrap so that it can receive SNMP v1 traps

[20-Sep-2011 08:52:11] <hmp> whats the difference between maintenance and decommissioned?

[20-Sep-2011 08:52:54] <hmp> found it, sorry

[20-Sep-2011 08:55:17] <hmp> doesnt seem to be working though

[20-Sep-2011 08:56:22] <hmp> why does zenoss still ping/snmp the device and throws an event if it's in maintenance

[20-Sep-2011 09:21:37] <jmp242> hmp - it's by design

[20-Sep-2011 09:21:47] <jmp242> the maintenance is just for the alerting rule matching

[20-Sep-2011 09:23:26] <hmp> so i have to do it manualy in the evt transform?

[20-Sep-2011 09:23:50] <hmp> manually*

[20-Sep-2011 09:26:44] <jmp242> If you don't want events yes

[20-Sep-2011 09:26:52] <jmp242> or you can use decommissioned

[20-Sep-2011 09:29:11] <hmp> tried, still throwing events, lol

[20-Sep-2011 09:45:24] <SDuensin> Greetings.

[20-Sep-2011 09:47:19] <dhopp> hmp: decommissioned should not throw events.  However, the daemons may not pick up that change right away.  Have you restarted daemons?

[20-Sep-2011 10:36:35] <SDuensin> Anybody got a minute for some idiot-level SNMP help?

[20-Sep-2011 10:38:01] <dhopp> SDuensin:  what's the question?

[20-Sep-2011 10:38:16] <SDuensin> How the hell does it work?  :-)  No, seriously...

[20-Sep-2011 10:38:31] <SDuensin> I built net-snmp and deployed it to a Windows machine.  I get this:  http://pastebin.com/wiwmTF9k

[20-Sep-2011 10:40:00] <dhopp> SDuensin:  that looks like you are missing a bunch of modules that would look up the OID mappings…it shouldn't actually stop snmp from working (although you might not be able to use 'system' at the end of your command)

[20-Sep-2011 10:40:23] <SDuensin> So it's compiled wrong?  Or it can't find the libs?

[20-Sep-2011 10:40:30] <dhopp> SDuensin:  at the end you have "Timeout: No Response from localhost"…did you start snmp?  Is it configured to listen for the community 'public' from 127.0.0.1 ?

[20-Sep-2011 10:41:22] <SDuensin> It's pretty much "out-of-the-box".

[20-Sep-2011 10:41:30] <dhopp> SDuensin:  it can't find the files..It's been a long time since I've compiled net-snmp by hand so I don't remember how you get those (might be a separate download)

[20-Sep-2011 10:42:37] <SDuensin> dhopp, ok, I'll read more.  I just wasn't sure if I was using it wrong, if it was something obvious I skipped, or if I screwed the install.  Seems it's the last one.

[20-Sep-2011 10:42:38] <dhopp> also..you might want to try snmpwalk -v2c -c public localhost

[20-Sep-2011 10:43:25] <SDuensin> About the same.

[20-Sep-2011 10:43:26] <SDuensin> :-/

[20-Sep-2011 10:43:51] <illsci> morning

[20-Sep-2011 10:45:55] <SDuensin> Thanks, dhopp

[20-Sep-2011 10:45:58] <SDuensin> Hi illsci

[20-Sep-2011 10:47:01] <illsci> I have a question about this warning im getting

[20-Sep-2011 10:47:06] <illsci> for all my hosts...

[20-Sep-2011 10:47:26] <illsci> Unable to read processes on device 10.252.171.69; Timeout on device

[20-Sep-2011 10:47:34] <illsci> what is causing that.... My snmpd.conf view has .1

[20-Sep-2011 10:47:45] <illsci> so I think it should be able to see everything there is to see... right?

[20-Sep-2011 10:47:50] <dhopp> illsci:  what snmp version are you using?

[20-Sep-2011 10:47:52] <dhopp> 1 or 2c?

[20-Sep-2011 10:47:55] <dhopp> in zenoss

[20-Sep-2011 10:48:00] <illsci> you mean the protocol?

[20-Sep-2011 10:48:02] <illsci> 1

[20-Sep-2011 10:48:07] <dhopp> change it to 2c

[20-Sep-2011 10:48:13] <dhopp> and see if that clears it up

[20-Sep-2011 10:48:52] <LW_> Good morning,  I've got Zenoss 3.1.0 running and have had no problems for a few months, then last week a process on a host started reporting down no matter what we would do, it would re-model and re-find the process, but then would constantly report the process as down, now starting this morning, the same thing has occured with six different hosts, and I'm wondering if anyone has seen anything like that occuring before, and if they could

[20-Sep-2011 10:49:18] <illsci> http://dpaste.com/617759/

[20-Sep-2011 10:49:38] <dhopp> illsci:  no you need to tell Zenoss to use 2c

[20-Sep-2011 10:49:40] <illsci> hey this thats my snmpd.conf

[20-Sep-2011 10:49:45] <dhopp> It's a configuration property in Zenoss

[20-Sep-2011 10:49:50] <illsci> ahhh ok

[20-Sep-2011 10:49:54] <dhopp> it defaults to 1

[20-Sep-2011 10:50:10] <illsci> is that a Zproperty

[20-Sep-2011 10:50:13] <dhopp> yes

[20-Sep-2011 10:50:38] <illsci> im trying to remember how to navigate to that stuff...

[20-Sep-2011 10:50:47] <dhopp> ZSnmpVer

[20-Sep-2011 10:50:56] <dhopp> illsci:  what version of zenoss? 3?

[20-Sep-2011 10:51:08] <illsci> asvanced -> collectors -> localhost something?

[20-Sep-2011 10:51:11] <illsci> 3.2.0

[20-Sep-2011 10:51:11] <dhopp> no

[20-Sep-2011 10:51:25] <dhopp> illsci:  Click Infrastructure

[20-Sep-2011 10:51:59] <dhopp> then drill down to the device class or device you want to change (or you can change it at the "Device Classes" level to affect all hosts).

[20-Sep-2011 10:52:06] <dhopp> Click the Blue Details button

[20-Sep-2011 10:52:10] <dhopp> Click Configuration Properties

[20-Sep-2011 10:52:18] <dhopp> Find ZSnmpVer and change it to 2c

[20-Sep-2011 10:52:26] <dhopp> then restart zenperfsnmp

[20-Sep-2011 10:52:39] <dhopp> might need to restart zenhub first..not sure

[20-Sep-2011 10:53:24] <illsci> I'll just restart the whole thing

[20-Sep-2011 10:53:31] <dhopp> that works too

[20-Sep-2011 10:54:25] <illsci> can you restart individual pieces through the web interface?

[20-Sep-2011 10:54:52] <dhopp> you can if you aren't using distributed collectors under Advanced -> Daemons

[20-Sep-2011 11:02:00] <illsci> hmm

[20-Sep-2011 11:03:05] <illsci> ok I just deleted all the events

[20-Sep-2011 11:03:15] <illsci> Let me see if they come back

[20-Sep-2011 11:08:36] <illsci> Is there anyway I can test it out and see if it sees processes?

[20-Sep-2011 11:09:08] <illsci> The snmpwalk command through the interface doesn't show much different..

[20-Sep-2011 11:09:14] <illsci> its only asking for system

[20-Sep-2011 11:09:16] <illsci> though...

[20-Sep-2011 11:10:29] <smaguire> illsci: when are you seeing that error, during modelling?

[20-Sep-2011 11:12:36] <illsci> well I have a bunch of hosts already added

[20-Sep-2011 11:12:44] <illsci> So I'm waiting for them to be remodeled I think

[20-Sep-2011 11:13:30] <illsci> They are all blank... no blue

[20-Sep-2011 11:13:35] <illsci> Just nothing so far...

[20-Sep-2011 11:13:57] <illsci> Anyway to see if any of this is working....

[20-Sep-2011 11:14:01] <illsci> other than wait?

[20-Sep-2011 11:15:10] <smaguire> snmpwalk -v2c -c YOURCOMMUNITY localhost prTable ?

[20-Sep-2011 11:15:25] <smaguire> without the '?'

[20-Sep-2011 11:16:14] <smaguire> I'm not 100% sure that's the table it uses but it seems like a likely candidate

[20-Sep-2011 11:16:16] <illsci> UCD-SNMP-MIB::prTable = No Such Object available on this agent at this OID

[20-Sep-2011 11:17:58] <smaguire> "If no proc directives are defined, then walking the prTable will fail (noSuchObject)."

[20-Sep-2011 11:18:11] <smaguire> that's from here: "http://www.net-snmp.org/docs/man/snmpd.conf.html"

[20-Sep-2011 11:18:43] <illsci> ok...

[20-Sep-2011 11:18:46] <illsci> let me hook this up

[20-Sep-2011 11:21:57] <smaguire> hmm, I've had a quick look and it seems like for that table you need to specifically enable snmp monitoring of each process

[20-Sep-2011 11:23:44] <illsci> I'm surprised this isn't documented more...

[20-Sep-2011 11:24:00] <illsci> I bought a zenoss book and it had a suggested snmpd.conf in it and it makes no mention of  this stuff

[20-Sep-2011 11:24:31] <smaguire> your snmpd.conf is almost exactly the same as mine

[20-Sep-2011 11:24:45] <smaguire> if you add something this at the end of the file on a new line:

[20-Sep-2011 11:24:51] <smaguire> proc httpd

[20-Sep-2011 11:25:04] <illsci> Do you do any process monitoring?

[20-Sep-2011 11:25:27] <smaguire> then restart snmp, you'll see an extry in the prTable for httpd (assuming it's running)

[20-Sep-2011 11:26:07] <illsci> Are there any suggested configs for all the different processes...

[20-Sep-2011 11:26:16] <smaguire> we don't at the moment, we're still on Nagios for that, but we're looking at moving it to zenoss soon

[20-Sep-2011 11:26:28] <smaguire> so I was interested in seeing how it worked

[20-Sep-2011 11:26:30] <illsci> I was hoping to manage this stuff through zenoss... but I guess if you have to configure snmpd.conf per each process...

[20-Sep-2011 11:26:39] <smaguire> maybe some other people in here will have more experience

[20-Sep-2011 11:26:57] <smaguire> well, you could try using a command plugin instead of snmp

[20-Sep-2011 11:27:06] <illsci> I wish there was a way to enable access to all the processes and then monitor the ones you care about in zenoss

[20-Sep-2011 11:27:07] <Simon4> you don't need to do that

[20-Sep-2011 11:28:13] <Simon4> walk 1.3.6.1.2.1.25.4.2.1

[20-Sep-2011 11:28:27] <Simon4> that's where zenoss gets it's list of running processes, and how it monitors whether a process is running

[20-Sep-2011 11:28:49] <illsci> yeah thats working

[20-Sep-2011 11:28:57] <illsci> hmmm ok well how do I ensure that's working

[20-Sep-2011 11:29:13] <illsci> Do I have to add processes to be monitored first

[20-Sep-2011 11:29:28] <illsci> yeah let me do that..

[20-Sep-2011 11:29:42] <Simon4> so the trick is to go to Infrastructure -> Processes

[20-Sep-2011 11:29:53] <Simon4> create the process you want monitored

[20-Sep-2011 11:30:00] <Simon4> then remodel the device that should have that process

[20-Sep-2011 11:30:02] <Simon4> it should appear

[20-Sep-2011 11:30:09] <illsci> How do you remodel the devices

[20-Sep-2011 11:30:26] <Simon4> navigate to a device, then choose "remodel" from the gear menu at the bottom left

[20-Sep-2011 11:31:19] <smaguire> i stand corrected

[20-Sep-2011 11:31:29] <illsci> the only options are Bind Templates, Reset Bindings, Clear Geocode Cache...

[20-Sep-2011 11:32:00] * Simon4 wonders if they moved it in 3.2

[20-Sep-2011 11:32:44] <smaguire> yeah, that does work, cheers Simon4

[20-Sep-2011 11:32:51] <Simon4> cool

[20-Sep-2011 11:33:00] <illsci> Now I just need to find out how to remodel the devices...

[20-Sep-2011 11:33:16] <Simon4> illsci: worst case as the zenoss user on the cli run zenmodeler run -d devicename

[20-Sep-2011 11:33:25] <Simon4> but there should be a way in the GUI, I just can't remember where it lives

[20-Sep-2011 11:34:35] <smaguire> on 2.5 it's in the dropdown when viewing the device under Manage

[20-Sep-2011 11:34:56] <Simon4> 3.1 it was in the gear menu

[20-Sep-2011 11:35:15] * smaguire fears upgrading and losing everything

[20-Sep-2011 11:36:48] <dhopp> it's still in the gear menu on 3.2

[20-Sep-2011 11:36:48] <illsci> supposedly its in the same spot but I'm not seeing it for some reason...

[20-Sep-2011 11:37:01] <dhopp> illsci: are you on a device or device class?

[20-Sep-2011 11:37:04] <dhopp> you need to be on a device

[20-Sep-2011 11:38:34] <tvincent> Survey: Upgrade to 3.2.1 when it is released or just skip to 4.0.2?

[20-Sep-2011 11:39:12] <dhopp> tvincent:  are you an enterprise customer?

[20-Sep-2011 11:39:21] <tvincent> yeap

[20-Sep-2011 11:39:35] <Hackman238> tvincent: Go to 4

[20-Sep-2011 11:39:38] <dhopp> what I know of version 4 is that it's a huge upgrade…many things fixed

[20-Sep-2011 11:39:59] <illsci> ok when I drill down to a device I can see it

[20-Sep-2011 11:40:04] <Hackman238> tvincent: If you're at >3.1 it shouldnt hurt too much to upgrade

[20-Sep-2011 11:40:09] <dhopp> tvincent:  you will lose your events though because the event system is completely different..unless Hackman238 knows a way around that

[20-Sep-2011 11:40:33] <Hackman238> tvincent: Theres a script to convert old events db to the new schema, but it is not pretty.

[20-Sep-2011 11:40:44] <Hackman238> tvincent: Thats the only caveat

[20-Sep-2011 11:41:18] <tvincent> Will it fix a problem where I have 80 servers brought up and ZenOSS decides to model all 5000 processes.This has happened to me a couple of times and its really a process fix on my end working around it. Brings things to a dead crawl.

[20-Sep-2011 11:41:28] <illsci> even after remodeling the device it still doesn't show up as Blue

[20-Sep-2011 11:41:33] <illsci> Is that normal?

[20-Sep-2011 11:41:35] <illsci> no events?

[20-Sep-2011 11:41:59] <Hackman238> tvincent: Zenmodel in v4 spreads the work out

[20-Sep-2011 11:42:18] <Hackman238> tvincent: Though 5K processes on 80 servers seems like...a lot.

[20-Sep-2011 11:42:30] <tvincent> Yes it is.

[20-Sep-2011 11:42:37] <dhopp> it's only 600 process per server :-P

[20-Sep-2011 11:42:50] <tvincent> dhopp: No, 5000 processes PER server.

[20-Sep-2011 11:42:52] <dhopp> tvincent:  are you trying to monitor every process possible?

[20-Sep-2011 11:42:58] <hmp> dhopp: yea, that was the problem, thx

[20-Sep-2011 11:43:04] <Hackman238> tvincent: You'll want to spreact thopse across logical collectors

[20-Sep-2011 11:43:06] <tvincent> dhopp: know its the app.

[20-Sep-2011 11:43:32] * dhopp scrolling back to figure out what hmp's problem was

[20-Sep-2011 11:43:52] <tvincent> Support gave me some dmd to delete all the ip services  from those services. Ended up fixing the problem.

[20-Sep-2011 11:44:16] <dhopp> hmp: ahh..decommissioned devices still throwing events.  I've ran into that before :-)

[20-Sep-2011 11:44:24] <Hackman238> tvincent: Yeah if you dont want to monitor those processes I'd remove the process modeler fromt hat class

[20-Sep-2011 11:44:31] <illsci> When I add processes the bottom right menu says Loading forever...

[20-Sep-2011 11:44:38] <illsci> Is that a bug?

[20-Sep-2011 11:44:54] <tvincent> Hackman238: Whats the real advantage of 4.0.2? I know it has a revamped message bus, queing system, etc.

[20-Sep-2011 11:44:56] <dhopp> illsci:  what browser are you using?

[20-Sep-2011 11:45:04] <illsci> Chrome...

[20-Sep-2011 11:45:07] <illsci> let me try on firefox

[20-Sep-2011 11:45:11] <dhopp> use firefox

[20-Sep-2011 11:45:12] <illsci> the last time it behaved exactly the same

[20-Sep-2011 11:45:15] <dhopp> and in chrome..delete your cookies

[20-Sep-2011 11:45:40] <Hackman238> tvincent: graphs use rrdcache, faster. rrdcache also doubles the number of dp disk io can handle. New event system, updated and useful event console

[20-Sep-2011 11:45:59] <Hackman238> tvincent: New alerting system cleaner, more flexible, extensible

[20-Sep-2011 11:46:25] <Hackman238> tvincent: Zope faster & fewwer bottle necks at hub processes

[20-Sep-2011 11:46:50] <Hackman238> tvincent: Daemons ore resilient to connection failure and headvy load

[20-Sep-2011 11:46:52] <tvincent> have you upgraded?

[20-Sep-2011 11:46:56] <Hackman238> tvincent: Trying to

[20-Sep-2011 11:47:09] <Hackman238> tvincent: Out upgrade is much harder

[20-Sep-2011 11:47:20] <Hackman238> tvincent: We run v2.5.2 over 12 instances

[20-Sep-2011 11:47:33] <Hackman238> tvincent: Each instance with several collectors

[20-Sep-2011 11:48:08] <Hackman238> tvincent: All auomated with custom scripts, CRM processes, ect

[20-Sep-2011 11:48:30] <tvincent> nice

[20-Sep-2011 11:48:52] <Hackman238> tvincent: Its been a f*ng nightmare. Smaller places with less tie-ins and newer versions will ahve no problem. V3.1->v4.0.2 is smooth as baileys

[20-Sep-2011 11:48:57] <tvincent> We are starting to dip our feet in the water with more adventurous stuff.

[20-Sep-2011 11:49:10] <dhopp> Hackman238:  with distributed collectors in core how does zenmodeler work?  Does it run on each collector and only model the devices that collector is responsible for? or does it run from the central hub?

[20-Sep-2011 11:49:13] <tvincent> automatic events, piping event console to a irc channell.

[20-Sep-2011 11:49:26] <Hackman238> tvincent: Yep- we do that also. Its a great idea

[20-Sep-2011 11:49:48] <Hackman238> dhopp: In Ent each logical collector has a modeler that does its device set

[20-Sep-2011 11:49:56] <Hackman238> dhopp: in core it depends on the setup

[20-Sep-2011 11:50:08] <tvincent> right now I am trying to hack in more automatic integration with JIRA.

[20-Sep-2011 11:51:00] <Hackman238> tvincent: If you're going to upgrade, upgrade now and dev that later\

[20-Sep-2011 11:51:12] <tvincent> Has the API changed alot?

[20-Sep-2011 11:51:22] <Hackman238> tvincent: Lots of facilitiy has been stabndardized, updated, extended, in v4 comparred to v3

[20-Sep-2011 11:51:35] <Hackman238> tvincent: Yes and no. Its way cleaned up

[20-Sep-2011 11:51:50] <Hackman238> tvincent: The idea of how things work is the same, but its more standard

[20-Sep-2011 11:52:02] <dhopp> Hackman238:  ok..I have 2 physical collectors, each with 3 logical collectors (daemons defined in daemons.txt).  If I wanted to schedule zenmodeler to run from cron could I just scheduled col1_zenmodeler, col2_zenmodeler, etc?

[20-Sep-2011 11:52:28] <dhopp> col1,col2 being the logical collector running on a physical collector

[20-Sep-2011 11:52:36] <Hackman238> dhopp: If you want to cron them sure. They could run daemonized and spread it out too

[20-Sep-2011 11:53:09] <dhopp> Hackman238:  I read somewhere that it's can be better to cron the scheduled run rather then daemonized

[20-Sep-2011 11:53:38] <dhopp> I want whatever is easiest..but I don't want 6 zenmodeler processes trying to model all of my devices..that would be bad

[20-Sep-2011 11:53:58] <Hackman238> dhopp: true true

[20-Sep-2011 11:54:05] <Hackman238> dhopp: either way works

[20-Sep-2011 11:55:47] <JohnnyNOC> hello Hackman238... dhopp... everytone

[20-Sep-2011 11:55:51] <JohnnyNOC> everyone

[20-Sep-2011 11:56:08] <Hackman238> JohnnyNOC: Helo

[20-Sep-2011 11:56:24] <JohnnyNOC> Hackman238 how ya living

[20-Sep-2011 11:57:07] <JohnnyNOC> default answer: "in living color"

[20-Sep-2011 11:57:27] <Simon4> "by breathing"

[20-Sep-2011 11:57:43] <JohnnyNOC> Simon4 that certainly helps!

[20-Sep-2011 11:57:56] <Hackman238> JohnnyNOC: Living hell!

[20-Sep-2011 11:58:00] <Hackman238> JohnnyNOC: Just kidding

[20-Sep-2011 11:58:19] <JohnnyNOC> hahaha.. that reminds me

[20-Sep-2011 11:58:29] <JohnnyNOC> i saw something online that i thought you might like..  because i thought they were german

[20-Sep-2011 11:58:33] <JohnnyNOC> google "cyberdance goth party"

[20-Sep-2011 11:58:48] <JohnnyNOC> i mean, i don't think you'll LIKE it but you may get a laugh or two

[20-Sep-2011 11:59:07] <Hackman238> JohnnyNOC: LOL!!!

[20-Sep-2011 11:59:54] <Simon4> woah

[20-Sep-2011 11:59:59] <Simon4> bad, wrong, badong

[20-Sep-2011 12:00:01] [disconnected at Tue Sep 20 12:00:01 2011]

[20-Sep-2011 12:00:01] [connected at Tue Sep 20 12:00:01 2011]

[20-Sep-2011 12:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[20-Sep-2011 12:01:04] <Hackman238> http://www.youtube.com/watch?v=fE8EMWxuZB0&feature=related

[20-Sep-2011 12:01:40] <JohnnyNOC> that is much cooler

[20-Sep-2011 12:03:32] * Simon4 plays at 11 on the home office amp

[20-Sep-2011 12:05:07] <Hackman238> http://www.youtube.com/watch?v=CX7ur4Ey5X4

[20-Sep-2011 12:05:22] <JohnnyNOC> i'll have to watch that at home

[20-Sep-2011 12:05:27] <JohnnyNOC> not sure it's SFmyW

[20-Sep-2011 12:05:31] <Hackman238> ^^Win. Theres an uncensored one somewhere

[20-Sep-2011 12:05:38] <Hackman238> JohnnyNOC: Its not

[20-Sep-2011 12:06:35] <JohnnyNOC> i do like the title

[20-Sep-2011 12:06:37] <JohnnyNOC> :]

[20-Sep-2011 12:06:45] <Hackman238> JohnnyNOC: Its a good  song anyway

[20-Sep-2011 12:07:05] <Hackman238> JohnnyNOC: Half german, half english

[20-Sep-2011 12:08:32] <dhopp> Hackman238:  so to be clear…in my setup if I assign 100 devices to col1, 100 devices to col2, 100 devices to col3, 100 devices to col4 and let col*_zenmodeler run as a daemon, each one of those will only model 100 devices?

[20-Sep-2011 12:08:58] <JohnnyNOC> dhopp that seems to be what i've experienced

[20-Sep-2011 12:12:20] <Hackman238> http://www.youtube.com/watch?v=8JjQGt7WjK0

[20-Sep-2011 12:12:27] <Hackman238> ^safe for work and Boss

[20-Sep-2011 12:13:18] <JohnnyNOC> hah

[20-Sep-2011 12:13:24] <JohnnyNOC> i hear metallica

[20-Sep-2011 12:13:40] <JohnnyNOC> on a violin

[20-Sep-2011 12:14:30] <Hackman238> JohnnyNOC: Well they are electric chello

[20-Sep-2011 12:15:21] <Hackman238> JohnnyNOC: Apocalyptica has some really great music

[20-Sep-2011 12:16:50] <JohnnyNOC> i feel like i've seen them on some tv show before though i'm not sure i could get into them

[20-Sep-2011 12:17:22] <Hackman238> JohnnyNOC: Yeah the youtube'd version not impressive.

[20-Sep-2011 12:17:30] <dhopp> Hackman238:  how granular do you have your RRDs? i.e. how long do you keep 5 minute samples?

[20-Sep-2011 12:18:23] <Hackman238> JohnnyNOC: http://www.youtube.com/watch?v=T0t-KRApDLs

[20-Sep-2011 12:18:31] <Hackman238> dhopp: Correct

[20-Sep-2011 12:18:44] <Hackman238> dhopp: They *should*

[20-Sep-2011 12:19:43] <dhopp> Hackman238:  ok..how about my next question :-P

[20-Sep-2011 12:20:11] <Hackman238> dhopp: One year

[20-Sep-2011 12:20:24] <Hackman238> dhopp: my rrd's are about 1.1MB

[20-Sep-2011 12:21:59] <dhopp> Hackman238:  ah…I was going to do 5 minutes for a month then 15 minutes for a year and 30 minutes for 2 years..but some people think we need more granular..so maybe I'll do 5 minutes for a year and 15 minutes for 2 years

[20-Sep-2011 12:22:25] <Hackman238> dhopp: It depends on the need for metrics really

[20-Sep-2011 12:24:36] <dhopp> well our clients big time is the last three months of the year (holiday) so we need to be able to do capacity planning…but for trending over 2 years I find it hard to imagine that I need to know that Nov 22 of 2009 at 2:15 the CPU was 20% and 2:20 it was 24% and 2:25 it was 22%..just that 15 minutes it averaged 22% (and then I can keep the max to get an idea of peak levels)

[20-Sep-2011 12:27:40] <nexex> dhopp, Hackman238 to change that -- do you modify the rrd command or is there a setting somewhere?

[20-Sep-2011 12:28:34] <Hackman238> http://eccentric.cx/misc/rrdcalc.html

[20-Sep-2011 12:28:46] <dhopp> nexex: there is a Default RRD Create command on each collector

[20-Sep-2011 12:29:00] <dhopp> then you can override it on a per datasource level

[20-Sep-2011 12:29:22] <Hackman238> nexex: you'll need to calculate a new create command, update each collector, delete all rrd's, restart daemons

[20-Sep-2011 12:31:59] <nexex> thanks

[20-Sep-2011 12:33:08] <luizzmizz> hi!

[20-Sep-2011 12:33:44] <luizzmizz> i've found a problem and i cannot solve it... hope someone can...

[20-Sep-2011 12:33:58] <luizzmizz> I can not access a Device Organizer (BridgeMIB)

[20-Sep-2011 12:34:10] <luizzmizz> and via zendmd when checking relations:

[20-Sep-2011 12:34:18] <luizzmizz> >>> for d in dmd.Devices.Network.Switch.getSubDevices():

[20-Sep-2011 12:34:24] <luizzmizz> ...   d.checkRelations(repair=True)

[20-Sep-2011 12:34:27] <luizzmizz> ...

[20-Sep-2011 12:34:28] <luizzmizz> 2011-09-20 18:31:33 WARNING ZenModel.DeviceOrganizer Stale deviceSearch record: /zport/dmd/Devices/Network/Switch/BridgeMIB/devices/10.128.104.253

[20-Sep-2011 12:42:44] <Hackman238> jmp242: Hey, you around?

[20-Sep-2011 12:58:38] <LW_> Hi,  I've got Zenoss 3.1.0 running and have had no problems for a few months, then last week a process on a host started reporting down no matter what we would do, it would re-model and re-find the process, but then would constantly report the process as down, now starting this morning, the same thing has occured with six different hosts, and I'm wondering if anyone has seen anything like that occuring before, and if they could point me i

[20-Sep-2011 13:01:56] <Hackman238> LW_: snmp down or ping down?

[20-Sep-2011 13:02:13] <jb> Hackman238: have you by chance ever used GPS coordinates for a location?

[20-Sep-2011 13:02:40] <Hackman238> jb: Yep- its not fun

[20-Sep-2011 13:02:58] <jb> I'm having trouble getting them to plot on the dashboard :/

[20-Sep-2011 13:03:02] <Hackman238> jb: Theres a script on the forums to do it automagically somewhere

[20-Sep-2011 13:03:15] <Hackman238> jb: Try clearning geocoded cache?

[20-Sep-2011 13:03:18] <jb> yeah

[20-Sep-2011 13:03:26] <Hackman238> jb: Does the map say geocoding?

[20-Sep-2011 13:03:29] <jb> i don't think there is a place in 3.x where you can view individual locations on a map..

[20-Sep-2011 13:03:30] <Hackman238> jb: Or nothing?

[20-Sep-2011 13:03:35] <jb> other than the dashboard

[20-Sep-2011 13:04:26] <Hackman238> jb: if you click the gear on the map portlet, what is the location set up?

[20-Sep-2011 13:04:31] <Hackman238> *set to?

[20-Sep-2011 13:04:38] <jb> hm one sec.

[20-Sep-2011 13:05:07] <jb> Base location is empty

[20-Sep-2011 13:05:44] <jb> "37.673706, -85.299581" is what I'm using for the address

[20-Sep-2011 13:06:03] <Hackman238> jb: set base locatio to /Locations/

[20-Sep-2011 13:06:06] <Hackman238> jb: save it

[20-Sep-2011 13:06:15] <Hackman238> jb: f5 the page

[20-Sep-2011 13:06:51] <jb> nothing still.. and if I clear the geocode cache, you can see the "Configuring 'X' Addresses" or whatever.. it only says 1

[20-Sep-2011 13:06:59] <jb> and that one location happens to have a real non-GPS address

[20-Sep-2011 13:07:13] <illsci> Hey are your nodes supposed to come up and be in the blue category

[20-Sep-2011 13:07:14] <illsci> ?

[20-Sep-2011 13:07:30] <Hackman238> illsci: thats a discovery evet

[20-Sep-2011 13:07:31] <jb> illsci: there will be a "Discovered Device" blue event for each device tha you add.

[20-Sep-2011 13:07:32] <Hackman238> *event

[20-Sep-2011 13:07:45] <Hackman238> jb: so the default location is set to /Locations ?

[20-Sep-2011 13:07:45] <illsci> ok

[20-Sep-2011 13:07:52] <illsci> let me remove them and readd them...

[20-Sep-2011 13:07:54] <jb> Hackman238: yep.

[20-Sep-2011 13:08:21] <jb> actually, it won't save the base location.

[20-Sep-2011 13:08:44] <Hackman238> jb: try setting to /Locations/Your location name

[20-Sep-2011 13:09:36] <jb> there it is.

[20-Sep-2011 13:09:49] <jb> sweet.

[20-Sep-2011 13:09:58] <Hackman238> jb: Glad its working

[20-Sep-2011 13:10:03] <jb> ok, so if I have a router at one location that is connected to a router at another location.. there should be a line, no?

[20-Sep-2011 13:10:15] <jb> and this functionality uses routing table maps?

[20-Sep-2011 13:10:21] <Hackman238> jb: If you're modeling routes and the routes make layer3 sense, yes

[20-Sep-2011 13:10:27] <Hackman238> jb: Yep

[20-Sep-2011 13:11:07] <jb> ok, let me populate the addresses for a few locations and see if this works.

[20-Sep-2011 13:11:08] <jb> thanks

[20-Sep-2011 13:19:59] <damjan> so I've added some mibs to my zenoss and now I get UncataloguedObjectException

[20-Sep-2011 13:20:47] <damjan> this is on Debian with zenoss installe from http://dev.zenoss.org/deb

[20-Sep-2011 13:21:48] <Hackman238> jb: NP

[20-Sep-2011 13:22:02] <Hackman238> damjan: Where?

[20-Sep-2011 13:22:36] <damjan> Hackman238: in the web interface, when I open the Advanced - MIBs page

[20-Sep-2011 13:22:52] <Hackman238> damjan: have you tried a reindex?

[20-Sep-2011 13:23:05] <damjan> how do I do that?

[20-Sep-2011 13:24:07] <Hackman238> damjan: su zenoss, `env python /opt/zenoss/Products/ZenUtils/zencatalog.py --forcereindex -v10`

[20-Sep-2011 13:24:15] <Hackman238> damjan: It will take a while

[20-Sep-2011 13:24:26] <Hackman238> damjan: also web ui device searchs wont work until its done

[20-Sep-2011 13:25:39] <humanix> hi there

[20-Sep-2011 13:25:45] <Hackman238> humanix: Hey

[20-Sep-2011 13:25:51] <humanix> hi Hackman238

[20-Sep-2011 13:25:52] <humanix>

[20-Sep-2011 13:26:05] <humanix> i have a doubt about os process monitoring in zenoss 2.5.x

[20-Sep-2011 13:26:38] <jb> hrm this locations stuff does not work very well.

[20-Sep-2011 13:26:41] <damjan> Hackman238: will try it now

[20-Sep-2011 13:27:13] <Hackman238> jb: Nope, not really. in v4 its improved

[20-Sep-2011 13:27:16] <jb> if I have a base location set, and another device is directly connected to it, with valid routing tables.. nothing appears.

[20-Sep-2011 13:27:19] <Hackman238> humanix: Sure, whats up?

[20-Sep-2011 13:27:31] <jb> it's only showing the one device in the "base location

[20-Sep-2011 13:27:36] <humanix> i added zwinuser, password and disabled ignore

[20-Sep-2011 13:27:48] <Hackman238> jb: Oh thats layer3 shenanagains. You may need to add dummy routes to complete routing tables

[20-Sep-2011 13:28:16] <jb> even if the two locations have valid layer3 routes that are already modeled?

[20-Sep-2011 13:28:25] <humanix> it added winservices

[20-Sep-2011 13:28:30] <Hackman238> humanix: jb It should show all the locations at the level the default location is set to

[20-Sep-2011 13:28:31] <humanix> should i put on there the process name?

[20-Sep-2011 13:29:10] <Hackman238> jb: Well the routes might be correct, but there must be something missing to make it valid for layer3 without knowledge of layer2

[20-Sep-2011 13:29:21] <Hackman238> jb: layer3 topo isnt good.

[20-Sep-2011 13:29:35] <jb> i should think that if I set the base locations to /Locations, then I should at least see all dots on the map.

[20-Sep-2011 13:29:43] <Hackman238> humanix: Yes. It should populate a list as you type

[20-Sep-2011 13:30:01] <humanix> and it should show me any processes right?

[20-Sep-2011 13:30:21] <Hackman238> jb: You might want to create a master location like 'NY' or 'USA' or 'EARTH' and put everything in that

[20-Sep-2011 13:30:31] <damjan> Hackman238: btw, this is Zenoss 3.2.0

[20-Sep-2011 13:30:35] <Hackman238> jb: then use that as the default level.

[20-Sep-2011 13:30:39] <Hackman238> damjan: Yep

[20-Sep-2011 13:31:03] <Hackman238> jb: Note that dots under locations who have their own dot are combined in to the parent dot

[20-Sep-2011 13:32:01] <jb> right, i'm not seeing any dots at all when the base is set to /Locations

[20-Sep-2011 13:33:26] <Hackman238> jb: Yeah. You should, but as a work around int eh past I've just created a 'EARTH" location and put all the others under that. then use /locations/earth as the default

[20-Sep-2011 13:33:41] <jb> ahh

[20-Sep-2011 13:34:30] <Hackman238> humanix: It shoudl

[20-Sep-2011 13:34:40] <Hackman238> humanix: Thats not to say it will

[20-Sep-2011 13:34:53] <jb> ha, this is just full of bugs :/

[20-Sep-2011 13:34:55] <Hackman238> humanix: Process monitoring works much better under v4

[20-Sep-2011 13:35:13] <jb> i'm just going to wait on 4, I guess.

[20-Sep-2011 13:35:18] <Hackman238> jb: v4 improved that hugely

[20-Sep-2011 13:35:29] <Hackman238> jb: Its comming! Have you seen the new ZCA timeline?

[20-Sep-2011 13:35:33] <jb> yup.

[20-Sep-2011 13:35:38] <jb> hows that bigip pack coming?

[20-Sep-2011 13:35:38] <Hackman238> jb: Alrighty.

[20-Sep-2011 13:36:44] <Hackman238> jb: Haulted for the moment. We're massivly upgrading our backbone so I've been streched working with keeping zenoss up to date and debugging the ASR pack

[20-Sep-2011 13:36:55] <jb> ah

[20-Sep-2011 13:37:10] <Hackman238> jb: And what I say our backbone I mean worldwide- hundreds of ASR's

[20-Sep-2011 13:37:18] <jb> yeah

[20-Sep-2011 13:37:20] <jb> sounds fun

[20-Sep-2011 13:37:31] <Hackman238> jb: Yes

[20-Sep-2011 13:37:58] <Hackman238> jb: Rackspace great place to work

[20-Sep-2011 13:38:21] <jb>

[20-Sep-2011 13:38:39] <jb> yeah, I'd like to work for a large hosting shop.

[20-Sep-2011 13:39:25] <Hackman238> jb: We're always hiring.

[20-Sep-2011 13:42:22] * rocket notes he is qualified for janitor

[20-Sep-2011 13:44:36] <Hackman238> rocket: LOL!

[20-Sep-2011 13:44:54] <Hackman238> rocket: We need someone to rewrite CORE- thats where you'd get hired

[20-Sep-2011 13:45:13] <jb> your management software?

[20-Sep-2011 13:45:24] <humanix> i saw all, thanks Hackman238

[20-Sep-2011 13:45:27] <Hackman238> rocket: http://www.python.org/about/success/rackspace/

[20-Sep-2011 13:45:32] <Hackman238> humanix: NP

[20-Sep-2011 13:45:37] <Hackman238> jp: yep

[20-Sep-2011 13:45:54] <Hackman238> jb: Its great! *fake smile*

[20-Sep-2011 13:46:04] <jb> hah, i colo a box with a provider who's system is also called CORE.

[20-Sep-2011 13:46:14] <Hackman238> jb: LOL

[20-Sep-2011 13:46:39] <Hackman238> jb: We crack on it a lot here at RS, but the program has huge functionality.

[20-Sep-2011 13:49:54] <rocket> I prefer to claim "I know Nutink! I know Nutink!"

[20-Sep-2011 13:50:12] <Hackman238> rocket:

[20-Sep-2011 14:18:02] <damjan> Hackman238, so zencatalog.py --reindex didn't help

[20-Sep-2011 14:18:28] <damjan> Hackman238: I've found some forum post where the poster had the same issue, he solved it by renaming the objects in zope

[20-Sep-2011 14:19:29] <Hackman238> damjan: Thats bizarre. Can you post the trace and the post?

[20-Sep-2011 14:20:00] <damjan> message/61281

[20-Sep-2011 14:21:41] <damjan> Hackman238: what trace exactly do you need?

[20-Sep-2011 14:21:42] <Hackman238> damjan: ouch- not a good idea

[20-Sep-2011 14:22:04] <Hackman238> damjan: the web ui trace

[20-Sep-2011 14:22:17] <Hackman238> damjan: I'll try to give you some dmd foo to delete the object

[20-Sep-2011 14:23:28] <Hackman238> damjan: How many mibs do you see in the ZMI?

[20-Sep-2011 14:23:48] <damjan> I've added 4

[20-Sep-2011 14:24:14] <Hackman238> damjan: So if you go in the zmi you see only 4 entries for mibs?

[20-Sep-2011 14:24:31] <Hackman238> damjan: I ask because not all mibs load so nicely

[20-Sep-2011 14:24:48] <damjan> zmi is the zope management inteface right? not zenoss?

[20-Sep-2011 14:25:04] <damjan> here http://server:8080/zport/dmd/Mibs/mibs/manage_workspace I have 4 mibs that I've added

[20-Sep-2011 14:25:24] <damjan> I had to add  them one by one, since zenmib run didn't add them all at once

[20-Sep-2011 14:25:34] <Hackman238> damjan: If youtry to delete one, what happens?

[20-Sep-2011 14:25:46] <damjan> Hackman238: same error

[20-Sep-2011 14:25:57] <damjan> Hackman238: even with the base mib I get the error

[20-Sep-2011 14:26:12] <damjan> Hackman238: this is what I see in firebug:

[20-Sep-2011 14:26:23] <Hackman238> damjan: Thats odd. So if you rename does it error?

[20-Sep-2011 14:26:29] <damjan> a POST to mib_router with {"action":"MibRouter","method":"getOrganizerTree","data":["/zport/dmd/Mibs"],"type":"rpc","tid":4}

[20-Sep-2011 14:26:40] <damjan> I get back {"message": "UncataloguedObjectException ", "type": "exception"}

[20-Sep-2011 14:27:04] <damjan> Hackman238: when I rename them, let's say I add a .txt suffix the zenoss web ui then works

[20-Sep-2011 14:27:11] <Hackman238> damjan: Sounds like we'd need to use a bline _delOb on that. Not good.

[20-Sep-2011 14:27:22] <Hackman238> damjan: When you name it back does it work?

[20-Sep-2011 14:27:33] <damjan> Hackman238: let me try that

[20-Sep-2011 14:28:16] <Hackman238> damjan: If that works then its really creating a copy of the object as the new name and deleting the old one.

[20-Sep-2011 14:28:47] <damjan> renameing it back doesn't break the UI

[20-Sep-2011 14:28:48] <Hackman238> damjan: Which means a .delOb should work if you have a huge list. Since its only 4, rename by hand wont take long

[20-Sep-2011 14:29:28] <damjan> what's .delOb?

[20-Sep-2011 14:29:49] <Hackman238> damjan: Its a dmd method

[20-Sep-2011 14:30:00] <Hackman238> damjan: For programmatic operations

[20-Sep-2011 14:30:05] <damjan> is dmd the new Zope thing?

[20-Sep-2011 14:30:32] * damjan has had some Zope experience in the previous life

[20-Sep-2011 14:30:41] <Hackman238> damjan: dmd is a extended python interface to talk directly to zeo

[20-Sep-2011 14:30:48] <Hackman238> damjan: Ah gotcha

[20-Sep-2011 14:31:06] <rocket> do not use the zmi! .. almost always the zmi and zenoss is a bad idea

[20-Sep-2011 14:31:27] <damjan> rocket: it seems I had to here

[20-Sep-2011 14:32:12] <rocket> no .. I am serious .. it will break your zeo in strange and unusual ways

[20-Sep-2011 14:32:35] <Hackman238> damjan: Its true

[20-Sep-2011 14:32:47] <damjan> ok, I don't intend to

[20-Sep-2011 14:33:03] <rocket> trust me .. I know ..

[20-Sep-2011 14:34:58] <damjan> so, ok... finally I have my mibs in zenoss

[20-Sep-2011 14:35:24] <damjan> should I expect to see something when I click on a MIB in Advanced -> Mibs?

[20-Sep-2011 14:35:35] <damjan> because for most, I don't see anything

[20-Sep-2011 14:37:19] <Hackman238> damjan: Sounds like the mibs didnt import right

[20-Sep-2011 14:42:29] <damjan> well how can I see what went wrong?

[20-Sep-2011 14:46:07] <rocket> damjan I would remove the mibs from zmi and add them using the zenmib program

[20-Sep-2011 14:46:26] <damjan> rocket: I am trying to do it with the zenmib program

[20-Sep-2011 14:48:37] <damjan> btw, it's interesting that I can't delete them from the ZMI, I have to do the rename dance, and then delete them from the zenoss UI

[20-Sep-2011 14:48:52] <Hackman238> damjan: Tell me about it

[20-Sep-2011 14:49:07] <Hackman238> damjan: DMD is a huge savior

[20-Sep-2011 14:49:20] <Hackman238> damjan: you can get to it by su zenoss; zendmd

[20-Sep-2011 14:49:39] <Hackman238> damjan: your <Tab> <Tab> to list objects, methods and classes

[20-Sep-2011 14:50:54] <damjan> interesting

[20-Sep-2011 14:51:23] <damjan> Hackman238: so, I try to add 3 mibs (for now), is there some order required on the command line?

[20-Sep-2011 14:52:06] <Hackman238> damjan: Generally mibs work best when loaded in to $ZENHOME/share/mibs and running zenmib

[20-Sep-2011 14:54:40] <damjan> hm, so it seems there are errors in the mibs that zenoss doesn't like (though net-snmp didn't complain)

[20-Sep-2011 14:56:06] <Hackman238> damjan: In zenoss mibs dont do anything other that provide rtrap information

[20-Sep-2011 14:56:52] <damjan> well I was hoping to use it for traps

[20-Sep-2011 14:57:25] <damjan> Hackman238: but can't I use the mibs for monitoring too?

[20-Sep-2011 14:57:29] <jb> nope.

[20-Sep-2011 14:57:45] <jb> you need specific OID's for monitoring..

[20-Sep-2011 14:57:51] <jb> MIB's are just used to translate traps.

[20-Sep-2011 14:58:22] <Hackman238> jb: mibs are only for translation. templates built from oid's are for monitoring

[20-Sep-2011 14:58:39] <damjan> but responses need to be translated too

[20-Sep-2011 14:58:39] <jb> i know

[20-Sep-2011 14:58:52] <Hackman238> jb: Whoops! I meant that to damjan

[20-Sep-2011 14:59:12] <jb> damjan: example?

[20-Sep-2011 14:59:19] <Hackman238> damjan: Only for traps. Zenoss doesnt handle non numerical snmp data

[20-Sep-2011 14:59:38] <damjan> ah ok

[20-Sep-2011 14:59:47] <Hackman238> damjan: If you need non numerica snmp data you'll need to write a parser to convert it to a number representing the item

[20-Sep-2011 15:00:02] <Hackman238> damjan: You can also use a command datasource and do the same

[20-Sep-2011 15:13:39] <LW_> Hackman238: sorry for the delay, offsite meeting, but process down messages, but other processes are okay, it is specific processes

[20-Sep-2011 15:14:13] <Hackman238> LW_: I bet the clears ended up being lost

[20-Sep-2011 15:14:34] <Hackman238> LW_: I'd delete the processes and readd them

[20-Sep-2011 15:23:08] <LW_> Hackman238: tried that and made sure the events were gone, and when it remodels the host it finds the processes, re-adds them, and then reports that they are down

[20-Sep-2011 15:23:14] <jbk> hmm I have an snmp table where index is an item of interest, while index+1 is related to index (instead of being an independent entry)..

[20-Sep-2011 15:23:57] <jbk> while I know I can create a component w/ the modeler that creates instances (and stores the index/oid), I'd like to generate a graph for that component using data from index+1

[20-Sep-2011 15:24:45] <Hackman238> LW_: have you restarted zenhub; zenprocess and zenstatus by chance?

[20-Sep-2011 15:25:20] <Hackman238> LW_: su zenoss; zenprocess stop; zenstatus stop; zenhub stop; zenhub start; zenprocess start; zenstatus start; zenmodeler restart

[20-Sep-2011 15:25:28] <jbk> more specifically, got a netapp filer (the existing zenpack don't show much), so I've got a modeler that creates filesystems (by scanning the dfTable), but I want a graph showing the snapshot utilization for each filesystem which is a separate entry in the same table

[20-Sep-2011 15:25:33] <Hackman238> (will restart hub, process and status and modeler for good measure)

[20-Sep-2011 15:54:19] <LW_> Hackman238: restarted everything, I've got my setup in a H/A failover cluster so its a little harder then that, but I've tried twice already.

[20-Sep-2011 15:55:14] <Hackman238> LW_: Hum. I hate to say it, but you might need to delete that device and readd it

[20-Sep-2011 15:55:40] <LW_> at-least its fairly easy

[20-Sep-2011 15:55:43] * LW_ gives it a try

[20-Sep-2011 15:59:44] <LW_> so far so good, If its going to repeat the issue with the server i picked, it will take a minute

[20-Sep-2011 16:00:06] <Hackman238> LW_: Alrighty

[20-Sep-2011 16:05:21] <JohnnyNOC> anyone here using ibm xservers?

[20-Sep-2011 16:05:51] <tvincent> johnnyNOC: i have in the past.

[20-Sep-2011 16:06:16] <JohnnyNOC> tvincent was just wondering if there was anyone that has them and uses the IMM as there's a beta ZenPack for the IBM IMM now

[20-Sep-2011 16:12:25] <jmp242> I have an IBM xseries

[20-Sep-2011 16:12:30] <jmp242> but haven't tried the beta pack

[20-Sep-2011 16:12:36] <jmp242> I may try on release

[20-Sep-2011 16:18:02] <JohnnyNOC> jmp242 i'd appreciate hearing any feedback if/when you do

[20-Sep-2011 16:18:34] <JohnnyNOC> we were able to twist ibm's arm into creating the zenpack because they don't have a director agent for debian based distros

[20-Sep-2011 16:19:15] <JohnnyNOC> certainly still in beta so as i mentioned above i'd love to hear any feedback i could pass on

[20-Sep-2011 16:26:11] <LW_> Hackman238: worked for one host, didn't work for another it seems

[20-Sep-2011 16:44:49] <Hackman238> LW_: Weird

[20-Sep-2011 16:45:10] <Hackman238> LW_: Can you describe your failover syste,?

[20-Sep-2011 16:46:22] <LW_> Hackman238: DRBD+corosync... I think we have it figured out, looks like a single bit got flipped or something, our ignore arguemnts setting got changed somehow from off to on, we just realized it was only occuring on a single class of server

[20-Sep-2011 16:46:53] <Hackman238> LW_: ah gotcha

[20-Sep-2011 16:47:54] <LW_> which happened to run that process, which uses java, which all of the other java processes were alarming on those servers except that one

[20-Sep-2011 16:49:59] <LW_> what is really odd, is that two other servers that should have hit the same situation were un-affected, but I guess it all has to do with process modeling/checking order

[20-Sep-2011 17:21:05] kocolosk is now known as kocolosk|away

[20-Sep-2011 18:04:45] <lrmo> Does someone have a link for the Zenoss Core 3.1 download page?  I'm trying to upgrade to 3.2, but can't find 3.1 anywhere.

[20-Sep-2011 18:05:59] <lrmo> oh nvm, think I found it

[20-Sep-2011 19:07:14] fopsa is now known as forsberg

[20-Sep-2011 19:55:11] kocolosk|away is now known as kocolosk

[20-Sep-2011 21:26:15] bzed_ is now known as bzed

[20-Sep-2011 22:21:27] kocolosk is now known as kocolosk|away

[21-Sep-2011 00:00:01] [disconnected at Wed Sep 21 00:00:01 2011]

[21-Sep-2011 00:00:01] [connected at Wed Sep 21 00:00:01 2011]

[21-Sep-2011 00:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[21-Sep-2011 06:39:42] fOrsberg is now known as forsberg

[21-Sep-2011 06:40:16] forsberg is now known as fOrsberg

[21-Sep-2011 08:45:34] <Hackman238> Morning all!

[21-Sep-2011 08:51:10] <jmp242> morning

[21-Sep-2011 09:08:04] <Jane_Curry> Afternoon

[21-Sep-2011 09:08:17] <Jane_Curry> Quick sanity check, please...

[21-Sep-2011 09:08:54] <Jane_Curry> If I want to change the appearance of a standard Zenoss page, defined by a Core .pt file...

[21-Sep-2011 09:10:10] <Jane_Curry> can I simply create a copy of the pt file in my ZenPack, modify the bits I want, and that pt file should get picked up - in preference to the Core file?

[21-Sep-2011 09:18:53] <straterra> Hmm..is Zenoss fully IPv6 capable? As in...accessible over IPv6 and using IPv6 to monitor hosts?

[21-Sep-2011 09:19:19] <Hackman238> straterra: V4 is

[21-Sep-2011 09:19:28] <Hackman238> Jane_Curry: Hi Jane

[21-Sep-2011 09:19:40] <Hackman238> Jane_Curry: It should

[21-Sep-2011 09:20:22] <straterra> Hackman238: But V4 isn't out yet, is it?

[21-Sep-2011 09:20:52] <Hackman238> straterra: Soon.

[21-Sep-2011 09:20:58] <straterra> Fair enough

[21-Sep-2011 09:21:05] <Hackman238> straterra: This year

[21-Sep-2011 09:22:02] <straterra> That's soon enough for me

[21-Sep-2011 09:22:30] <straterra> Last time I checked, there wasn't an easy way to get

[21-Sep-2011 09:22:39] <straterra> Last time I checked, there wasn't an easy way to get Zenoss running on port 80...is this still correct?

[21-Sep-2011 09:23:21] <straterra> Not a big deal if it is..I can just toss a proxy infront of it

[21-Sep-2011 09:25:20] <Hackman238> straterra: Sure it it

[21-Sep-2011 09:25:22] <Hackman238> *it is

[21-Sep-2011 09:25:28] <enkrypt> I've got a Cisco ASA device, I add it, some snmp stuff is detected (such as its interfaces, and sysname/sysdescr) but it refuses to poll correctly, showing an orange event saying SNMP is down. when I check by hand with snmpwalk SNMP functions fine

[21-Sep-2011 09:25:37] <enkrypt> any idea what might be the problem?

[21-Sep-2011 09:25:44] <Hackman238> straterra: tweak zope.conf to point to 8080 and 80

[21-Sep-2011 09:26:09] <Hackman238> straterra: You can point to only 80 if you dont have existing items pointing to 8080

[21-Sep-2011 09:26:54] <Jane_Curry> Hackman238 - hmm. I have installed my ZenPack and it's still picking up the Core pt file

[21-Sep-2011 09:26:55] <Hackman238> straterra: vi $ZENHOME/etc/zope.conf and search 'address 8080'

[21-Sep-2011 09:26:56] <Simon4> enkrypt: change maxoidperrequest to be a low number like say 5 for those devices

[21-Sep-2011 09:27:04] <Hackman238> Jane_Curry: Did you restart zope?

[21-Sep-2011 09:27:16] <Hackman238> enkrypt: The device is being overwhelmed sounds like

[21-Sep-2011 09:27:18] <straterra> Hackman238: I might just leave it 8080 for standard sake and drop a proxy in front of it as I want to make it internet facing anyway

[21-Sep-2011 09:27:28] <Hackman238> enkrypt: What Simon4 said

[21-Sep-2011 09:27:53] <Hackman238> straterra: You definately want to setup SSL on that proxy if your zenoss will be internet facing

[21-Sep-2011 09:28:01] <straterra> Yup

[21-Sep-2011 09:28:04] <straterra> That's easy enough

[21-Sep-2011 09:28:17] <Hackman238> straterra: Zenoss, as like any monitoring tool, is a aweome point of entry and documentation to DoS and destroy the network

[21-Sep-2011 09:28:29] <Hackman238> straterra: Let me know if you need any help

[21-Sep-2011 09:28:32] <straterra> Will do

[21-Sep-2011 09:30:13] <enkrypt> Hackman238 & Simon4: ok trying to set it on 5

[21-Sep-2011 09:30:36] <Jane_Curry> The one I am modifying is down in ZenEvents/skins/zenevents - my copy is simply under my skins/<zenpack name> directory

[21-Sep-2011 09:30:48] <Jane_Curry> and yes, Ive done zopectl restart

[21-Sep-2011 09:31:00] <Hackman238> Jane_Curry: Gotcha. Let me look at something real quick

[21-Sep-2011 09:32:58] <Hackman238> Jane_Curry: Did specifiy the skin foo in the _init_.py?

[21-Sep-2011 09:33:02] <Hackman238> Jane_Curry: e.x.:

[21-Sep-2011 09:33:03] <Hackman238> skinsDir = os.path.join(os.path.dirname(__file__), 'skins')

[21-Sep-2011 09:33:05] <Hackman238> if os.path.isdir(skinsDir): registerDirectory(skinsDir, globals())

[21-Sep-2011 09:34:26] <enkrypt> I modified the maxoid setting for all Cisco ASA device class, I am working with remote collectors, how do I propagate this change immediately?

[21-Sep-2011 09:34:33] <enkrypt> do a collector update?

[21-Sep-2011 09:35:07] <Hackman238> enkrypt: You'll want to pushchange from each device

[21-Sep-2011 09:35:35] <enkrypt> did that. and didn't work, SNMP stays reported as down

[21-Sep-2011 09:36:00] <Hackman238> enkrypt: Okay I'd do a colelctor update, but only as an easy way to just restart all the collector daemons

[21-Sep-2011 09:36:18] <Hackman238> enkrypt: You also might need to clear that snmp down event as the clear could be lost

[21-Sep-2011 09:38:52] <Jane_Curry> Hackman238 - yup skinsDir registered

[21-Sep-2011 09:40:03] <Hackman238> Jane_Curry: I emailed you a pack of ours that replaces viewGraphs????.pt

[21-Sep-2011 09:42:31] <Jane_Curry> Hackman238 - thanks - is this going to cause me pain if I simply install it to see what is in it??

[21-Sep-2011 09:42:50] <Hackman238> Jane_Curry: I dont know...its converted for Avalon

[21-Sep-2011 09:43:00] <Hackman238> Jane_Curry: I dont think so- I thinnk it will just not work worst case

[21-Sep-2011 09:43:32] <Hackman238> Jane_Curry: It doesnt change much at all- but it will put our name on all your reportmail reports- maybe disassemble it first

[21-Sep-2011 09:43:54] <Jane_Curry> Sorry dissassemble?  Not sure how to unpack your egg....

[21-Sep-2011 09:44:41] <Hackman238> Jane_Curry: egg is just a zip

[21-Sep-2011 09:44:45] <Hackman238> Jane_Curry: Just unzip

[21-Sep-2011 09:45:20] <Hackman238> I just looked it over- it wont hurt anything

[21-Sep-2011 09:45:32] <enkrypt> Hackman238 & Simon4 works perfect now, thanks!

[21-Sep-2011 09:45:40] <Simon4> enkrypt: awesome

[21-Sep-2011 09:46:00] <Hackman238> It just updates the viewGraphsReportMail.py file

[21-Sep-2011 09:46:03] <Hackman238> enkrypt: Glad

[21-Sep-2011 09:46:18] <Hackman238> *.pt

[21-Sep-2011 09:47:07] <enkrypt> strange devices those Cisco ASA!

[21-Sep-2011 09:47:29] <Hackman238> enkrypt: Indeed

[21-Sep-2011 09:49:04] <jb> ive got to figure out how to AD integration working on this new 4 install.

[21-Sep-2011 09:49:09] <jb> it's never worked correctly :/

[21-Sep-2011 09:49:59] <Hackman238> jb: Whats not working?

[21-Sep-2011 09:50:41] <enkrypt> 4 install? zenoss 4.0 that is? what's it like? is it faster than 3, with respect to the GUI ?

[21-Sep-2011 09:51:10] <jb> oh, heh.. they yanked the 4.0 files already

[21-Sep-2011 09:51:34] <jb> Hackman238: it just doesn't work like it should.. and the documentation is poor.

[21-Sep-2011 09:51:38] <enkrypt> I'm holding out till 4.2

[21-Sep-2011 09:51:44] kocolosk|away is now known as kocolosk

[21-Sep-2011 09:51:49] <jb> Hackman238: i'd like to create AD security groups, and grant access by putting users into those groups

[21-Sep-2011 09:52:11] <jb> Hackman238: ie, users in "x" group have access to /Systems/Blah

[21-Sep-2011 09:52:17] <Hackman238> jb: Oh- not easy

[21-Sep-2011 09:52:37] <Hackman238> jb: The group mapping must be right, never an easy task

[21-Sep-2011 09:52:46] <jb> Hackman238: or, actually create the user in ZenOSS and assign managed devices, but use AD to authenticate them.

[21-Sep-2011 09:52:52] <Hackman238> enkrypt: Its coming! Trust me, its worth the wait!

[21-Sep-2011 09:53:23] <Hackman238> jb: Thats probably what you'll want to do. Group mapping can be done, but its always a hassle with AD.

[21-Sep-2011 09:54:34] <jb> Hackman28: I don't want just any AD user having access to zenoss, so if I can pre-define their accounts and access, that would be sufficient.

[21-Sep-2011 09:54:47] <enkrypt> Hackman238: I've got a 3.2.0 setup with 2300 monitored hosts (growing to 3000 by the end of this year), I would appreciate some extra performance

[21-Sep-2011 09:55:00] <jb> what kinds of devices?

[21-Sep-2011 09:55:35] <enkrypt> me? or Hackman238

[21-Sep-2011 09:55:41] <Hackman238> enkrypt: Expect double performance

[21-Sep-2011 09:56:08] <Hackman238> jb: Its doable, it seems to only work right with ldap though

[21-Sep-2011 09:56:11] <enkrypt> Hackman238: yeah I heard some rumours, the technical specs seem good. one thing that I didn't quite understand is why we stuck with RHEL5 instead of 6 ?

[21-Sep-2011 09:56:21] <Hackman238> jb: Can you instead do ldap against your ad?

[21-Sep-2011 09:56:33] <Hackman238> enkrypt: Theres good reasoning for this

[21-Sep-2011 09:56:46] <Hackman238> enkrypt: Partly to minimize the complication of migration

[21-Sep-2011 09:58:08] <Hackman238> enkrypt: Im sure there are technical reasons too

[21-Sep-2011 09:59:33] <enkrypt> ah ok

[21-Sep-2011 10:03:31] <rocket> Hackman238: I wouldnt necessarily set that expectation

[21-Sep-2011 10:04:14] <rocket> Hackman238: there are a few features of 4 that will not be as flexible in the core version, but it will still be very much faster

[21-Sep-2011 10:08:02] <jb> Hackman238: sure, LDAP is what I would use to talk to AD.

[21-Sep-2011 10:11:45] <Hackman238> jb: Do right now you've installed the ldap authenticator plugin?

[21-Sep-2011 10:22:42] <SDuensin> Morning.

[21-Sep-2011 10:27:05] <Hackman238> SDuensin: Morning

[21-Sep-2011 10:27:16] <SDuensin> Hackman238, dude, progress is being made!  :-)

[21-Sep-2011 10:27:34] <SDuensin> This has been a hell of a slog.

[21-Sep-2011 10:28:21] <Hackman238> SDuensin: Glad its moving in the right direction.

[21-Sep-2011 10:28:24] <Hackman238>

[21-Sep-2011 10:28:31] <SDuensin> Do I want to load all the MIBs that came with net-snmp into Zenoss?

[21-Sep-2011 10:31:18] <Hackman238> SDuensin: Personally I'd try to minimize the number of mibs as much as possible

[21-Sep-2011 10:31:37] <SDuensin> Got zero in there now.  Pretty minimal.

[21-Sep-2011 10:32:26] <SDuensin> Like everyone else, I just want Zenoss to tell me everything about everything.  :-)

[21-Sep-2011 10:34:58] <SDuensin> You should see my VNC solution.  I have noVNC (an HTML5 client) going over SSL to the "bouncer" proxy that then sends the connection over SSH to the target machine.  It's cool!

[21-Sep-2011 10:35:43] <jmp242> mmm

[21-Sep-2011 10:35:48] <jmp242> I just use logmein lol

[21-Sep-2011 10:36:00] <SDuensin> :-P

[21-Sep-2011 10:36:35] <SDuensin> I've got one-click deployment of our "Windows package" working, too.  FINALLY.  Man, what a freakin' mess that was.

[21-Sep-2011 10:37:27] <Hackman238> SDuensin: Very cool!

[21-Sep-2011 10:37:42] <SDuensin> Windows "security" == SUCK

[21-Sep-2011 10:38:05] <Hackman238> SDuensin: I'd be grateful if you could post the info on the madness that is windows security interoperability

[21-Sep-2011 10:38:16] <Hackman238> SDuensin: Indeed

[21-Sep-2011 10:38:47] <SDuensin> I want to.  Need to catch up here so I can continue having a job and then I want to write it all down.  Hopefully I can release the details of what I implemented.

[21-Sep-2011 10:38:56] <SDuensin> It is completely psychotic.

[21-Sep-2011 10:40:19] <Hackman238> SDuensin: We await the novel so we all might descend in to the madness XD

[21-Sep-2011 10:40:32] <Hackman238> SDuensin: Congrats on getting it to work- not an easy feat

[21-Sep-2011 10:40:50] <SDuensin> Really dude.  It's gonna make Stephen King's work look like freakin' children's tales.

[21-Sep-2011 10:41:00] <Hackman238> SDuensin: LOL!

[21-Sep-2011 10:45:56] <SDuensin> #1 thing heard in this office over the last month...   "F*CK!  _REALLY_?  Windows!"

[21-Sep-2011 10:46:54] <Hackman238> SDuensin: Sounds like deserves a reality TV show

[21-Sep-2011 10:46:58] <Hackman238> SDuensin: LOL

[21-Sep-2011 10:47:07] <SDuensin> It would have been better than Big Brother.

[21-Sep-2011 10:47:47] <SDuensin> What's really amusing is I got all this convoluted nonsense to work...  and then can't freakin' figure out some of the pretty menus in Zenoss.   (But we'll get to that later!)

[21-Sep-2011 10:48:33] <Hackman238> SDuensin: Personally I dont like the v3 UI either

[21-Sep-2011 10:49:07] <SDuensin> The whole templates thing makes zero sense to me.  But I haven't sat down with it in one window and the book in another yet either.

[21-Sep-2011 10:49:32] <Hackman238> SDuensin: The templating has been complicated in v3 unfortunately

[21-Sep-2011 10:49:44] <SDuensin> I'd say.

[21-Sep-2011 10:50:55] <Jane_Curry> Hackman238 - thanks for ZP - simple as you say - but I don't have a viewGraphReportMail.pt in my Core system - mebbe that's part of Enterprise

[21-Sep-2011 10:52:03] <Jane_Curry> Looks like you just have the skinsDir stuff in __init__.py as I have though

[21-Sep-2011 10:53:35] <Hackman238> Jane_Curry: Hum. Yeah its a pretty simple pack. It just overrides an existing template.

[21-Sep-2011 10:53:56] <Hackman238> Jane_Curry: Can you zip up what you have and email it for em to take a look?

[21-Sep-2011 11:17:10] <straterra> Hackman238: I came across an issue

[21-Sep-2011 11:17:12] <Jane_Curry> Hackman238 - email sent with ZenPack - any ideas appreciated

[21-Sep-2011 11:17:38] <straterra> The proxy is up and running..the login screen comes up, but my correct username/password don't work. Says it's invalid. I go to port 8080 and I can log in perfectly.

[21-Sep-2011 11:18:22] <Hackman238> Jane_Curry: I'll look soon. Anytime

[21-Sep-2011 11:18:25] <Hackman238> straterra: Whats up?

[21-Sep-2011 11:18:31] <straterra> I am changing the address that people are hitting, though

[21-Sep-2011 11:18:37] <Hackman238> straterra: Clear your cookies

[21-Sep-2011 11:18:44] <straterra> Alrighty

[21-Sep-2011 11:19:10] <Hackman238> straterra: If still fail, can you send me your apache config for your zenoss reverse proxy?

[21-Sep-2011 11:19:15] <straterra> It's nginx

[21-Sep-2011 11:19:18] <straterra> But I can

[21-Sep-2011 11:19:22] <Jane_Curry> Hackman238: tied up for an hour now, anyway

[21-Sep-2011 11:19:33] <Hackman238> straterra: Oh- IDK about nginx

[21-Sep-2011 11:19:42] <Hackman238> Jane_Curry: Alrighty

[21-Sep-2011 11:20:01] <straterra> It does basically the same thing

[21-Sep-2011 11:20:25] <straterra> Clearing cookies didn't work

[21-Sep-2011 11:20:43] <straterra> There isn't a 'hostname 'setting somewhere in zenoss, is there?

[21-Sep-2011 11:21:25] <Hackman238> straterra: Hostname?

[21-Sep-2011 11:21:48] <straterra> Yeah. Some web software have you set a hostname because they generate FQDN paths

[21-Sep-2011 11:22:02] <Hackman238> straterra: No, thats not a req in zenoss

[21-Sep-2011 11:22:16] <Hackman238> straterra: Ay referral should be accepted

[21-Sep-2011 11:22:17] <Hackman238> *any

[21-Sep-2011 11:24:24] <straterra> blogs/zenossblog/2010/10/28/tip-of-the-month-using-httpsssl-for-your-zenoss-ui looks very strange

[21-Sep-2011 11:24:33] <straterra> The rewrite rule for zope in particular

[21-Sep-2011 11:25:22] <Hackman238> straterra: Yep, that should work

[21-Sep-2011 11:25:32] <straterra> I'm not using that rewrite rule

[21-Sep-2011 11:26:37] <Hackman238> straterra: Yeah.

[21-Sep-2011 11:26:50] <straterra> That was the problem, I think

[21-Sep-2011 11:27:18] <straterra> It was

[21-Sep-2011 11:27:22] <straterra> Very very strange

[21-Sep-2011 11:27:34] <straterra> http://www.philipreichenberger.com/post/2333865615/zenoss-nginx-and-ssl gyi

[21-Sep-2011 11:27:36] <straterra> err, fyi

[21-Sep-2011 11:30:49] <Hackman238> straterra: Yeah the rewrite is needed for ssl config- I'm suprised a forwarding proxy didnt just work, however.

[21-Sep-2011 11:30:58] <Hackman238> straterra: Good to know

[21-Sep-2011 11:35:28] <straterra> Does http://alerts.marianinc.com load for you?

[21-Sep-2011 11:40:21] <straterra> Alright..hopefully this is the last issue. Our pages have links in them to delete the page, acknowledge, etc etc..but it's using the wrong host name. How do I change this?>

[21-Sep-2011 11:43:50] <Hackman238> straterra: Yep

[21-Sep-2011 11:44:26] <straterra> Aha, I think I found it.. zopeurl

[21-Sep-2011 11:46:02] <Hackman238> straterra: in zenactions.conf?

[21-Sep-2011 11:46:11] <straterra> Yeah

[21-Sep-2011 11:46:13] <Hackman238> straterra: That'll just change the zope url for links

[21-Sep-2011 11:46:21] <Hackman238> straterra: In alert emails, etc

[21-Sep-2011 11:46:23] <straterra> Which is want I want, isn't it?

[21-Sep-2011 11:46:27] <straterra> Yeah..that's what I want

[21-Sep-2011 11:46:39] <Hackman238> straterra: Yes, but it wouldnt have stopped proxy referal

[21-Sep-2011 11:46:43] <Hackman238> straterra: Correct

[21-Sep-2011 11:46:58] <straterra> No no no, of course not. I'm changing it so that my alerts have the proxied host name

[21-Sep-2011 11:47:09] <Hackman238> straterra: Yeah that'll do it

[21-Sep-2011 11:47:10] <Hackman238>

[21-Sep-2011 11:47:18] <straterra> And zenjobs

[21-Sep-2011 11:58:42] <jb> erm, anyone happen to know where x11 stores window sizes for applications?

[21-Sep-2011 11:58:54] <jb> i accidentally made an app's window very tiny, and now everytime I launch it, it's tiny.

[21-Sep-2011 12:00:01] [disconnected at Wed Sep 21 12:00:01 2011]

[21-Sep-2011 12:00:01] [connected at Wed Sep 21 12:00:01 2011]

[21-Sep-2011 12:00:12] [zenoss-logger (logger bot) has joined #zenoss]

[21-Sep-2011 12:25:42] <Hackman238> Wow...

[21-Sep-2011 12:25:44] <Hackman238> http://www.youtube.com/watch?v=iJ4T9CQA0UM&feature=player_embedded

[21-Sep-2011 12:25:48] <Hackman238> ^^Must watch this

[21-Sep-2011 12:26:14] <Hackman238> "Guy on a buffalo!"

[21-Sep-2011 12:32:54] <pulseman> hi everyone

[21-Sep-2011 12:33:21] <pulseman> there is a brazillian on the house??

[21-Sep-2011 12:54:36] bigegor_ is now known as bigegor

[21-Sep-2011 13:02:13] <wayne___> what was the mechanism for telling zenoss that one event class clears a different event class?

[21-Sep-2011 13:24:14] <straterra> Hackman238: you around?

[21-Sep-2011 13:25:17] <Hackman238> straterra: Hey

[21-Sep-2011 13:25:24] <Hackman238> wayne___: I'm not sure I understand

[21-Sep-2011 13:25:55] <Hackman238> BTW all....much like BeetleJuice, you need to say my name to summon my attention

[21-Sep-2011 13:26:04] <straterra> Small question..I'm getting an alert that there's an error with my transform/mapping for /Unknown..but..I log in Zenoss and I can't even seem to FIND the transform

[21-Sep-2011 13:26:52] <Hackman238> straterra: Interesting

[21-Sep-2011 13:27:19] <straterra> I can normally find it by going to Events -> Event Classes -> Unknown

[21-Sep-2011 13:27:20] <straterra> right?

[21-Sep-2011 13:27:31] <Hackman238> straterra: Yep

[21-Sep-2011 13:27:37] <Hackman238> straterra: Lets try a report

[21-Sep-2011 13:27:41] <Hackman238> straterra: docs/DOC-3428

[21-Sep-2011 13:28:00] <Hackman238> straterra: download the egg and rename it so the py2.4 is py2.6. Install the egg and restart Zope.

[21-Sep-2011 13:28:16] <Hackman238> straterra: In Reports there will be a new report for finding all transforms everywhere

[21-Sep-2011 13:29:00] <straterra> I haven't installed an egg since the first version of zenoss heh

[21-Sep-2011 13:29:06] <straterra> Can I do it from the web interface these days?

[21-Sep-2011 13:29:17] <Hackman238> straterra: Yep

[21-Sep-2011 13:29:26] <Hackman238> But you'll need to rename it still

[21-Sep-2011 13:29:39] <Hackman238> straterra: And you'll still need to restart zope

[21-Sep-2011 13:30:52] <straterra> Ok, I installed it from the web interface, but its not in the list. Does the zope restart make it visible?

[21-Sep-2011 13:33:29] <straterra> Erhm..the egg won't install

[21-Sep-2011 13:34:35] <JohnnyNOC> anyone have ideas on how i might be able to import an rrd from nagios and have zenoss display it?

[21-Sep-2011 13:34:49] <JohnnyNOC> i haven't had any luck just dropping in the rrd and creating the graph with a fake datapoint

[21-Sep-2011 13:35:25] <JohnnyNOC> oh, duh, let me look through some zenrender logs

[21-Sep-2011 13:36:20] <Hackman238> straterra: Yep

[21-Sep-2011 13:36:42] <straterra> I did a restart of zope and it's not in the egg list

[21-Sep-2011 13:36:43] <Hackman238> straterra: did you unzip the zip? IDK why but all the packs are zipes

[21-Sep-2011 13:36:50] <Hackman238> *zips

[21-Sep-2011 13:36:51] <straterra> No

[21-Sep-2011 13:36:56] <straterra> duh

[21-Sep-2011 13:37:01] <straterra> I'm stupid..I didn't even notice that

[21-Sep-2011 13:37:18] <Hackman238> straterra: download it, unzip, and there will be an egg. Rename the egg py2.4->py2.6 and try that

[21-Sep-2011 13:37:22] <Hackman238> straterra: Sorry

[21-Sep-2011 13:37:26] <straterra> Not your fault heh

[21-Sep-2011 13:37:58] <Hackman238> JohnnyNOC: Sure

[21-Sep-2011 13:38:05] <straterra> There it is

[21-Sep-2011 13:38:29] <Hackman238> JohnnyNOC: All you need to do is make sure the DS and ranges match those used in zenoss if you intend to put more data in

[21-Sep-2011 13:38:34] <Hackman238> straterra:

[21-Sep-2011 13:38:41] <Hackman238> straterra: In reports run tha transform report

[21-Sep-2011 13:39:08] <straterra> I don't see it under reports

[21-Sep-2011 13:39:15] <JohnnyNOC> Hackman238 *nod*

[21-Sep-2011 13:39:17] <straterra> there is it

[21-Sep-2011 13:39:20] <straterra> it is, rather

[21-Sep-2011 13:39:22] <JohnnyNOC> problem was the name of the datasource wasn't ds0

[21-Sep-2011 13:40:02] <Hackman238> JohnnyNOC: Gotcha. RRD dump it, modify the xml to match, rrdtool restore, bam!

[21-Sep-2011 13:40:07] <JohnnyNOC> yea

[21-Sep-2011 13:40:10] <JohnnyNOC> it's working now

[21-Sep-2011 13:40:12] <JohnnyNOC> thank you Sir

[21-Sep-2011 13:40:14] <Hackman238> JohnnyNOC: Oh nice

[21-Sep-2011 13:40:27] <straterra> Hackman238: I don't see it in the report

[21-Sep-2011 13:41:59] <Hackman238> straterra: Darn...that means tis stuck in a cache somewhere or theres shenanagains going on

[21-Sep-2011 13:42:12] <Hackman238> straterra: zenhub is reporting the bad transform or zenactions?

[21-Sep-2011 13:42:13] <straterra> Maybe I should restart zenoss completely?

[21-Sep-2011 13:42:32] <Hackman238> straterra: Not yet, not if we dont have to

[21-Sep-2011 13:42:33] <straterra> I've no idea..I've just been getting pages on it

[21-Sep-2011 13:42:45] <Hackman238> straterra: Can you PM me the message?

[21-Sep-2011 13:42:58] <straterra> I'll paste bin it

[21-Sep-2011 13:43:02] <Hackman238> straterra: TY

[21-Sep-2011 13:43:18] <straterra> http://pastebin.com/Wfx4Tume

[21-Sep-2011 13:43:58] <wayne___> Hackman238: I remember there being a way to configure an event type so that Zenoss knows it clears another event type...but I don't remember what that mechanism was

[21-Sep-2011 13:48:12] <Hackman238> straterra: su zenoss; zendmd

[21-Sep-2011 13:48:19] <Hackman238> straterra: run 'dmd.Events.Unknown.transform'

[21-Sep-2011 13:48:24] <Hackman238> straterra: Wahts it return?

[21-Sep-2011 13:48:41] <Hackman238> wayne___: Clear events?

[21-Sep-2011 13:48:56] <Hackman238> wayne___: Or transforms to tie to differnt event classes together

[21-Sep-2011 13:49:05] <Hackman238> s/to/two/

[21-Sep-2011 13:49:27] <straterra> 'if evt.oid == \'1.3.6.1.4.1.6876.4.1.301\':\n evt.summary = "vSphere Alarm"\n evt.eventClass = "/HW/ESX"\n evt.message = "vSphere Hardware Alarm"'

[21-Sep-2011 13:50:11] <Hackman238> straterra: copy that and save it somewhere

[21-Sep-2011 13:50:18] <wayne___> Hackman238: transforms to tie them together.  I've event.severity = 0 in the transform for the clear event, but I don't remember how to make it clear the other

[21-Sep-2011 13:50:26] <straterra> Hackman238: done

[21-Sep-2011 13:50:53] <Hackman238> straterra: dmd.Events.Unknown.transform = ''

[21-Sep-2011 13:50:56] <Hackman238> straterra: commit()

[21-Sep-2011 13:51:08] <Hackman238> straterra: zenhub restart; zenactions restart

[21-Sep-2011 13:51:17] <Hackman238> straterra: See if the pages continue

[21-Sep-2011 13:51:45] <straterra> the zenhub restart command failed

[21-Sep-2011 13:52:03] <Hackman238> straterra: do zenhub run now -v10

[21-Sep-2011 13:52:08] <Hackman238> straterra: pastebin the trace

[21-Sep-2011 13:52:11] <Hackman238> plz

[21-Sep-2011 13:52:19] <straterra> it doesn't like anything after the zenhub command

[21-Sep-2011 13:52:24] <straterra> invalid syntax

[21-Sep-2011 13:52:34] <Hackman238> straterra: What? Thats weird

[21-Sep-2011 13:52:43] <straterra> one sec and you can see

[21-Sep-2011 13:52:51] <Hackman238> straterra: As user zenoss?

[21-Sep-2011 13:53:00] <straterra> yes..but still in that console

[21-Sep-2011 13:53:02] <Hackman238> wayne___: It depends on how you can tie the together

[21-Sep-2011 13:53:16] <Hackman238> wayne___: s/the/them/

[21-Sep-2011 13:53:24] <straterra> Lemme try outside of the console

[21-Sep-2011 13:53:29] <straterra> it's restarting

[21-Sep-2011 13:53:30] <Hackman238> wayne___: What events are you trying to correlate?

[21-Sep-2011 13:53:38] <Hackman238> straterra: Alrighty

[21-Sep-2011 13:53:46] <straterra> restarted

[21-Sep-2011 13:53:59] <straterra> Now we play the waiting game

[21-Sep-2011 13:54:16] <Hackman238> straterra:

[21-Sep-2011 13:57:45] <straterra> Hackman238: Got an alert

[21-Sep-2011 13:58:04] <straterra> It's on line 0 now instead of 4..but otherwise same alert

[21-Sep-2011 13:58:45] <wayne___> Hackman238: I created an ospfNbrStateChange event class.  When the state changes to full it should clear several other events - from ospfNbrStateChange down events (same class) to ospfIfAuthFailure

[21-Sep-2011 14:01:39] <Hackman238> straterra: Interesting. Seems almost like its getting a malformed value in the OID

[21-Sep-2011 14:01:58] <Hackman238> wayne___: Ah gotcha. That could take some serious time to craft

[21-Sep-2011 14:02:00] <straterra> Problem with line 0: if evt.oid == '1.3.6.1.4.1.6876.4.1.301':

[21-Sep-2011 14:02:12] <straterra> But..if I understand correctly, we deleted the transform entirely

[21-Sep-2011 14:03:19] <Hackman238> straterra: Thats correct.

[21-Sep-2011 14:03:50] <straterra> I can restart the bloody thing

[21-Sep-2011 14:03:52] <Hackman238> straterra: Oh line 0, not 1.3.6.1.4.1.6876.0

[21-Sep-2011 14:04:01] <Hackman238> straterra: Thats weird

[21-Sep-2011 14:04:04] <straterra> Indeed

[21-Sep-2011 14:04:05] <Hackman238> straterra: Sounds like you might need to

[21-Sep-2011 14:04:22] <Hackman238> straterra: Seems like the transform is lodged somewhere

[21-Sep-2011 14:04:23] <straterra> I need to move zenoss to a newer server

[21-Sep-2011 14:04:32] <Hackman238> straterra: v2.5.2?

[21-Sep-2011 14:04:33] <straterra> Some screens take a good minute

[21-Sep-2011 14:04:41] <Hackman238> straterra: Oh damn

[21-Sep-2011 14:04:42] <straterra> No

[21-Sep-2011 14:04:57] <Hackman238> straterra: Thats got to be painful

[21-Sep-2011 14:04:59] <straterra> 3.1.0

[21-Sep-2011 14:05:01] <straterra> Sometimes, yes

[21-Sep-2011 14:06:31] <straterra> I just restarted zenoss...and got a page

[21-Sep-2011 14:06:47] <straterra> I think its an old one though

[21-Sep-2011 14:07:31] <straterra> omfg..I'm an idiot

[21-Sep-2011 14:07:35] <Hackman238> straterra: >

[21-Sep-2011 14:07:37] <Hackman238> straterra: ?

[21-Sep-2011 14:07:50] <straterra> Time: 2011/08/04 16:04:45.000

[21-Sep-2011 14:07:57] <straterra> It's an ancient alert

[21-Sep-2011 14:08:03] <Hackman238> straterra: *face palm*

[21-Sep-2011 14:08:15] <Hackman238>

[21-Sep-2011 14:08:25] <straterra> so..how do I load that transform back?

[21-Sep-2011 14:08:46] <straterra> Actually

[21-Sep-2011 14:08:49] <Hackman238> straterra: I'd use the UI and paste it back

[21-Sep-2011 14:09:03] <Hackman238> straterra: you could zendmd and set it back the same way we blanked it

[21-Sep-2011 14:09:07] <straterra> In the events dashboard...it's showing today as the last seen date

[21-Sep-2011 14:09:16] <Hackman238> straterra: try to use the UI when possible though

[21-Sep-2011 14:09:23] <Hackman238> straterra: Hum

[21-Sep-2011 14:09:34] <straterra> I'ma try to delete it and see if it comes back

[21-Sep-2011 14:09:45] <Hackman238> straterra: Alrighty

[21-Sep-2011 14:10:33] <straterra> Shouldn't acknowledging the alert shut it up though?

[21-Sep-2011 14:11:11] <Hackman238> straterra: Yes and no. Yes except when its a zenoss exception in some cases

[21-Sep-2011 14:11:16] <straterra> Gotcha

[21-Sep-2011 14:12:07] <straterra> Well..it didn't come back

[21-Sep-2011 14:12:43] <Hackman238> straterra: Nice. Let it go for a bit.

[21-Sep-2011 14:13:49] <jmp242> Hackman238: is there an issue with Decomissioned status not working with 3.2.0?

[21-Sep-2011 14:17:25] <straterra> Hmm...how would I go about auto-clearing issues when they don't exist anymore? In our current Nagios, we get a page when something goes down and a page when it comes up and the problem auto-clears when it comes up

[21-Sep-2011 14:17:45] <tvincent> anyone know how to hard code the interface speed for a device within zenoss?

[21-Sep-2011 14:22:39] <Hackman238> jmp242: Not that I know of- is it not working/

[21-Sep-2011 14:23:13] <Hackman238> straterra: Most items should clear with clear messages already. Traps you'll need to write transforms to handle

[21-Sep-2011 14:23:30] <Hackman238> tvincent: Yes, but it'll get reset by the interface modeller

[21-Sep-2011 14:23:41] <straterra> I don't use traps except for a single ESX trap..so thats fine

[21-Sep-2011 14:23:51] <jmp242> I've seen it posted a few times like here: message/61465#61465

[21-Sep-2011 14:23:52] <Hackman238> tvincent: d=dmd.Devices.findDevice('your device')

[21-Sep-2011 14:23:54] <straterra> Except that's not the behaviour I'm seeing..I have a lot of old stuff in my events console

[21-Sep-2011 14:24:11] <Hackman238> tvincent: i=d.os.interfaces._getOb('interface name')

[21-Sep-2011 14:24:19] <straterra> To get a clear page, I have to delete the object in the event console

[21-Sep-2011 14:24:24] <Hackman238> tvincent: i.speed = any int

[21-Sep-2011 14:24:38] <Hackman238> straterra: For what type of event

[21-Sep-2011 14:24:39] <Hackman238> ?

[21-Sep-2011 14:25:26] <straterra> I have quite a few for process not running

[21-Sep-2011 14:25:43] <straterra> They're running now and the last seen is like 4 months ago..but still there

[21-Sep-2011 14:25:59] <Hackman238> straterra: Process clears almost never work in v3.1

[21-Sep-2011 14:26:06] <Hackman238> straterra: They are working in v4.0.2

[21-Sep-2011 14:26:14] <Hackman238> straterra: Not certain about v3.2

[21-Sep-2011 14:26:20] <straterra> Fantastic

[21-Sep-2011 14:26:28] <dhopp> straterra: what is the severity?  error?

[21-Sep-2011 14:26:36] <straterra> Some warning, some error

[21-Sep-2011 14:26:53] <Hackman238> straterra: The best way, I find, to handle stale events it to write some sql foo to close all events that are x days old

[21-Sep-2011 14:27:07] <Hackman238> straterra: x being the number of days before it could be stale realistically

[21-Sep-2011 14:27:21] <straterra> Ah, good idea

[21-Sep-2011 14:27:34] <Hackman238> straterra: It also make sure problems arent ignored by techs who just ack events

[21-Sep-2011 14:27:58] <Hackman238> straterra: Cause it'll harrass them again in x days since they didnt fix *that* maxed out port or down server, etc

[21-Sep-2011 14:28:10] <straterra> Good point

[21-Sep-2011 14:28:45] <Hackman238> straterra: huge problem with people using zenoss....people ack stuff and ignore it. Put an end t that crap!

[21-Sep-2011 14:33:18] <tvincent> Hackman238 thanks!

[21-Sep-2011 14:33:33] <Hackman238> tvincent: NP

[21-Sep-2011 14:37:00] <jbk> for a legend value... can you use a python expression? i want to display the value 'min([expr1, expr2])'

[21-Sep-2011 14:37:39] <jbk> for the filesystem utilization thresholds, instead of a straight percentage, i'm using min(85%, 100gb) so that for very large filesystems, it's not complaining because there's 'only' 500gb free (for example)

[21-Sep-2011 14:37:56] <jbk> and I'd like the legend value to be the actual number

[21-Sep-2011 14:38:24] <Hackman238> jbk: I do not think so. Tales, RPN and RRD only

[21-Sep-2011 14:39:04] <Hackman238> jbk: You could, however, precalculate the value and dump it in to a cProperty. you can then refer to the cProp in the RRD via ${here/cPropName}

[21-Sep-2011 15:10:25] <straterra> Hackman238: I've deleted a device..but I'm still getting pages about it being down

[21-Sep-2011 15:14:49] <Hackman238> straterra: restart zenping

[21-Sep-2011 15:14:55] <Hackman238> straterra: See if that stops it

[21-Sep-2011 15:27:30] <Hackman238> straterra: Win? Fail?

[21-Sep-2011 15:29:41] <straterra> One sec, I'm about to

[21-Sep-2011 15:29:47] <straterra> Getting hammered with phone calls

[21-Sep-2011 15:30:56] <Hackman238> straterra: Alrighty.

[21-Sep-2011 15:31:10] <straterra> I restarted it..so..now I wait

[21-Sep-2011 15:31:29] <Hackman238> straterra: Alrigty

[21-Sep-2011 15:35:11] <straterra> Worked..thanks

[21-Sep-2011 15:36:30] <Hackman238> straterra: NP

[21-Sep-2011 15:36:53] <straterra> Acknowledging some of these alerts isn't stopping my email pages though

[21-Sep-2011 15:37:01] <straterra> Maybe I need to add a filter to the alert

[21-Sep-2011 15:38:08] <straterra> Aha, only if event state = New

[21-Sep-2011 15:39:37] <Hackman238> straterra: yep. make sure the alert rule has ==New

[21-Sep-2011 15:40:15] <Hackman238> straterra: It will stop most. Acked alerts caused by zenoss probs sometimes still happen in v2.5.2. No idea how.

[21-Sep-2011 15:51:48] <straterra> This new interface messes with my head...How can I change the location of a device now?

[21-Sep-2011 15:52:32] <straterra> Oh wait..drag and drop

[21-Sep-2011 15:53:01] <Hackman238> straterra: infrastructure page. Drag and drop in to locations...maybe. Not 100% on that one

[21-Sep-2011 15:53:23] <straterra> Yeah, drag and drop

[21-Sep-2011 15:53:28] <straterra> I was looking for a drop down

[21-Sep-2011 15:53:56] <Hackman238> straterra: Yeah...drag and drop is silly

[21-Sep-2011 16:05:31] <straterra> Hackman238: Looks like disk usage doesn't auto-clear either

[21-Sep-2011 16:05:52] <straterra> Oops, spoke too soon

[21-Sep-2011 16:06:39] <Hackman238> straterra: Disks should

[21-Sep-2011 16:35:54] <SDuensin> Might I just mention that WMI sucks?  Thanks.

[21-Sep-2011 16:52:42] <nexex> If I added a device by IP address initially, and now have it using the DNS name, is it safe to rename the directory for the device in zenoss/perf/ to match? Would need to shutdown zenoss first I would guess...

[21-Sep-2011 16:55:15] <dhopp> nexex: use the "rename device" in the gear menu in Zenoss

[21-Sep-2011 16:58:21] <nexex> dhopp: ah, even better thanks

[21-Sep-2011 17:09:27] kocolosk is now known as kocolosk|away

[21-Sep-2011 17:18:08] <Hackman238> Have to run.

[21-Sep-2011 17:18:10] <Hackman238> Later all!

[21-Sep-2011 17:19:10] <nonsenso> phew.

[21-Sep-2011 17:20:49] <SDuensin> Has anyone ever made WMI work?

[21-Sep-2011 17:21:42] <dhopp> I had it working at my last job..but I don't think we had UAC on

[21-Sep-2011 17:21:53] <dhopp> or had it turned way down

[21-Sep-2011 17:21:58] <SDuensin> I can't hit an XP box either though.  :-(

[21-Sep-2011 17:22:21] <dhopp> are there any firewalls?

[21-Sep-2011 17:22:31] <SDuensin> Not anymore.

[21-Sep-2011 17:24:45] <SDuensin> Using the same credentials, I can query a W7 box from the XP box using wbemtest.  Attempting the same from Linux to the W7 box errors.

[21-Sep-2011 17:27:27] <dhopp> WMI required DCOM ports and by default the DCOM port range is stupid (pretty everything above 1024 could be used)

[21-Sep-2011 17:27:30] <dhopp> with what error?

[21-Sep-2011 17:27:45] <SDuensin> Failed to bind to uuid 4d9f4ab8-7d1c-11cf-861e-0020af6e7c57 - NT_STATUS_END_OF_FILE

[21-Sep-2011 17:29:18] <SDuensin> What about the machine that is issuing the query?

[21-Sep-2011 17:31:22] <dhopp> I don't think it requires anything special.  Are you using an admin account?

[21-Sep-2011 17:31:35] <SDuensin> Domain admin, yea.

[21-Sep-2011 17:31:45] <SDuensin> Tried it with the Linux firewall off.  No change.

[21-Sep-2011 17:32:24] <dhopp> is the NT_STATUS_END_OF_FILE the last error?  Or is there a NT_STATUS_ACCESS_DENIED after that?

[21-Sep-2011 17:33:07] <SDuensin> Entire error:  http://pastebin.com/rTPpJF4u

[21-Sep-2011 17:33:27] <dhopp> NTSTATUS: NT_STATUS_ACCESS_DENIED - Access denied

[21-Sep-2011 17:33:49] <dhopp> when you specified the user did you use just 'user' or 'DOMAIN\user'

[21-Sep-2011 17:33:50] <dhopp> ?

[21-Sep-2011 17:33:57] <SDuensin> Yes.  domain/user%pass

[21-Sep-2011 17:34:05] <SDuensin> Also tried domain\\user%pass

[21-Sep-2011 17:34:11] <SDuensin> And domain\\user --password pass

[21-Sep-2011 17:35:39] <dhopp> and you are sure the password is right?  Does the password have any special characters (like a '$')?

[21-Sep-2011 17:38:01] <SDuensin> Yes, it's correct.  No, no specials.

[21-Sep-2011 17:38:17] <dhopp> docs/DOC-2520

[21-Sep-2011 17:38:38] <dhopp> according to that the Access denied can be either because 1) you are using a limited account or 2) your username/password actually is wrong

[21-Sep-2011 17:38:53] <dhopp> but it's probably not an exhaustive list

[21-Sep-2011 17:39:15] <SDuensin> yup

[21-Sep-2011 17:39:20] <SDuensin> Read that.  :-/

[21-Sep-2011 17:40:58] <dhopp> can you try to create a local user on that server and put it in the local administrators group?

[21-Sep-2011 17:41:16] <SDuensin> If GP will let me.

[21-Sep-2011 17:41:55] <dhopp> oh…your GP might be restricting wmi as well..is it maybe requiring a kerberos connection?

[21-Sep-2011 17:44:52] <SDuensin> Grr.  Gonna have to resume this tomorrow.  Local account didn't help matters.

[21-Sep-2011 17:45:57] <SDuensin> dhopp, thanks for the help.

[21-Sep-2011 17:46:02] <dhopp> no problem

[21-Sep-2011 17:56:33] <Gonzo1> Hi all, I'm currently running Zenoss on CentOS 5 and I'm running across some minor issues with SSL Cert checks. Everytime I run: /opt/zenoss/zenoss/ZenPacks/ZenPacks.Iwillfearnoevil.Domain-1.0.6-py2.6.egg/ZenPacks/Iwillfearnoevil/Domain/libexec/check_certs_Nagios.sh -x 60 -n -s 1.2.3.4 -p 443 I get errors: unable to load certificate

[21-Sep-2011 17:56:33] <Gonzo1> 20623:error:0906D06C:PEM routines:PEM_read_bio:no start line:pem_lib.c:647:Expecting: TRUSTED CERTIFICATE

[21-Sep-2011 17:56:33] <Gonzo1> unable to load certificate

[21-Sep-2011 17:56:33] <Gonzo1> 20626:error:0906D06C:PEM routines:PEM_read_bio:no start line:pem_lib.c:647:Expecting: TRUSTED CERTIFICATE

[21-Sep-2011 17:56:33] <Gonzo1> unable to load certificate

[21-Sep-2011 17:56:34] <Gonzo1> 20629:error:0906D06C:PEM routines:PEM_read_bio:no start line:pem_lib.c:647:Expecting: TRUSTED CERTIFICATE

[21-Sep-2011 17:56:34] <Gonzo1> unable to load certificate

[21-Sep-2011 17:56:35] <Gonzo1> 20633:error:0906D06C:PEM routines:PEM_read_bio:no start line:pem_lib.c:647:Expecting: TRUSTED CERTIFICATE

[21-Sep-2011 17:56:35] <Gonzo1> Status Expired -2455826 days ago.| 192.168.245.61 443 Expired  Days=--2455826

[21-Sep-2011 17:56:47] <Gonzo1> Does anyone know what I can do to resolve this?

[21-Sep-2011 18:50:17] <hikenboot> anyone able to tell me if there is a solaris client that can send traps to my linux server?

[21-Sep-2011 19:01:35] <hikenboot> ok forget snmp anyone know of an agent for solaris to that will work with zenoss

[21-Sep-2011 19:19:37] <jbk> well solaris10 should include net-snmp (granted it's a bit old)

[21-Sep-2011 19:20:40] <jbk> the service is application/management/sma (don't ask why they used that name)

[21-Sep-2011 19:21:52] <jbk> that'll grab the basic stuff

[21-Sep-2011 19:22:05] <jbk> and you can also turn on the fault management module

[21-Sep-2011 19:22:16] <jbk> so any fma events can be queried

[21-Sep-2011 19:22:22] <jbk> or send traps

[21-Sep-2011 19:50:22] kocolosk|away is now known as kocolosk

[21-Sep-2011 21:01:32] johnf1912 is now known as johnf1911

[21-Sep-2011 22:30:47] <bender183> sma is the old solaris 10 way for controlling services

[21-Sep-2011 22:35:23] <jbk> no

[21-Sep-2011 22:35:32] <jbk> sma = system management agent

[21-Sep-2011 22:35:37] <jbk> smf = new way to manage services

[21-Sep-2011 22:35:43] <jbk> sysv init = old way to manage services

[21-Sep-2011 22:35:52] <jbk> sma is essentially an old version of net-snmp

[21-Sep-2011 22:36:21] <jbk> i think in what will become solaris 11, they might have made things a bit more sane

[21-Sep-2011 22:36:41] <jbk> there was an old snmp agent that was junk that was still around for compatability (sea IIRC)

[21-Sep-2011 23:02:09] kocolosk is now known as kocolosk|away

[22-Sep-2011 00:00:01] [disconnected at Thu Sep 22 00:00:01 2011]

[22-Sep-2011 00:00:01] [connected at Thu Sep 22 00:00:01 2011]

[22-Sep-2011 00:00:01] [disconnected at Thu Sep 22 00:00:01 2011]

[22-Sep-2011 00:00:02] [connected at Thu Sep 22 00:00:02 2011]

[22-Sep-2011 00:00:16] [zenoss-logger (logger bot) has joined #zenoss]

[22-Sep-2011 03:54:41] <Jane_Curry>  hikenboot: get thye net-snmp agent for Solaris - I have used that quite happily with Solaris in the past

[22-Sep-2011 04:12:57] <Madmucho> Good morning

[22-Sep-2011 04:24:39] <Madmucho> Can someone please help me with zenoss ? I have problem when deleting monitored server.

[22-Sep-2011 04:28:34] <Madmucho> Dont know when im at good adress here :-/

[22-Sep-2011 04:50:52] <pokui> what port does zope listen on when running zenoss 3.x?

[22-Sep-2011 04:54:21] <Simon4> 8080

[22-Sep-2011 04:56:11] <pokui> Simon4: hmm.. is there a way to talk to zope directly rather than zenoss? (trying to fix my rewrite rules)

[22-Sep-2011 04:56:31] <Simon4> http://hostname:8080/manage may do it

[22-Sep-2011 04:56:54] <pokui> thanks. let me try that.

[22-Sep-2011 06:49:32] <enkrypt> using zendmd, how do I get all devices based on certain criteria, for example which collector they are polled by, or by device class?

[22-Sep-2011 06:54:52] <enkrypt> when I try to change the collector for a large amount of devices via the GUI I get this error: StaleResultsException Search results do not match

[22-Sep-2011 09:10:30] <jmp242> hey, Hackman238 - I'm thinking about a blog post on 3.2 suggesting waiting for 3.2.1 or making sure you've installed the patches. This info hasn't really been spread to the community

[22-Sep-2011 09:11:18] <Hackman238> jmp242: Thats a good idea

[22-Sep-2011 09:11:44] <jmp242> Can I also promote your Zenpacks as a ZCA beta?

[22-Sep-2011 09:12:28] <Hackman238> jmp242: Sure

[22-Sep-2011 09:14:00] <enkrypt> whenever I try to use rmatte's windows cpu zenpack, even for only 1 host, my zenhub goes insane

[22-Sep-2011 09:14:50] <enkrypt> Zenoss enterprise 3.2.0 + the 3 patches. and the newest version of rmatte's zenpack, the Simple variant, version 1.6

[22-Sep-2011 09:14:56] <enkrypt> bc and snmpwalk commands are present

[22-Sep-2011 09:15:01] <enkrypt> when I execute the command by hand it works fine

[22-Sep-2011 09:15:15] <enkrypt> but still, as soon as I activate the zenpack, all other daemons start to complain that my zenhub isn't running

[22-Sep-2011 09:18:48] <jmp242> ZCA Update: community/zca/blog/2011/09/22/zca-update-for-september-22-2011--zenoss-32-warning-zenpacks

[22-Sep-2011 09:35:14] <Hackman238> enkrypt: Which pack by ryan?

[22-Sep-2011 09:36:01] <Hackman238> enkrypt: sounds like you're going to want to enable config caching and pcache in zenohub

[22-Sep-2011 09:36:49] <uniplexed> hi there

[22-Sep-2011 09:37:12] <humanix> damn is me

[22-Sep-2011 09:37:18] <humanix> i was connected already

[22-Sep-2011 09:37:19] <humanix> hi there

[22-Sep-2011 09:37:22] <humanix> rmatte, hi

[22-Sep-2011 09:37:24] <humanix> are you there?

[22-Sep-2011 09:37:46] <straterra> Hackman238: Good morning

[22-Sep-2011 09:38:31] <SDuensin> Greetings.

[22-Sep-2011 09:38:57] <Hackman238> Morning all

[22-Sep-2011 09:39:08] <humanix> monrinng Hackman238

[22-Sep-2011 09:39:11] <psyklone> howdy howdy

[22-Sep-2011 09:39:11] <humanix> xD

[22-Sep-2011 09:39:24] <humanix> i got a doubt again

[22-Sep-2011 09:39:37] <Hackman238> humanix: What about?

[22-Sep-2011 09:39:41] <humanix> is about counter32 and counter64 in interfaces

[22-Sep-2011 09:39:50] <humanix> i have seen that devices using v1 use counter32

[22-Sep-2011 09:40:21] <humanix> but this doesn't work for example on big exreme networks of cisco switches with giga or 10giga ifaces

[22-Sep-2011 09:40:29] <humanix> cause the value only fit on a base64

[22-Sep-2011 09:40:34] <enkrypt> Hackman238: Hackman238: this zenpack : docs/DOC-3570

[22-Sep-2011 09:40:59] <humanix> i was looking for the root of the problem, and i ve saw some devices using v2c that use a "mixed" template style

[22-Sep-2011 09:41:10] <Hackman238> humanix: Oh, thats no problem. You'll just want to force the use of v2c

[22-Sep-2011 09:41:12] <humanix> devices that have some interfaces on conter32 and others on 64

[22-Sep-2011 09:41:46] <humanix> the mixed ones complicate me, cause i dont know what is the root of configuration issue

[22-Sep-2011 09:42:14] <humanix> if devices use 2c, they can "know" what tipe of template do they ned per interface?

[22-Sep-2011 09:42:23] <humanix> and if they are using v1 they only use counter32?

[22-Sep-2011 09:42:28] <humanix> am i right?

[22-Sep-2011 09:42:35] <enkrypt> Hackman238: this is my current zenhub.conf: http://dpaste.com/618991/

[22-Sep-2011 09:43:30] <dhopp> enkrypt: workers 15?  That seems like a lot.  How many devices?

[22-Sep-2011 09:44:00] <Hackman238> enkrypt: 15 workers is too many

[22-Sep-2011 09:44:21] <Hackman238> enkrypt: Less you have 32 cores

[22-Sep-2011 09:45:10] <Hackman238> enkrypt: How many cores do you have?

[22-Sep-2011 09:45:16] <enkrypt> I'm running zenoss enterprise 3.2.0 + the 3 patches, 8 physical cores, 38gb ram, 2 distributed collectors

[22-Sep-2011 09:45:20] <Hackman238> enkrypt: Dont count hyper thread cores- they are worthless in python

[22-Sep-2011 09:45:44] <Hackman238> enkrypt: Gotcha

[22-Sep-2011 09:46:05] <Hackman238> enkrypt: in $ZENHOME/etc/zenhub.conf and for each collectorname-zenhub.conf:

[22-Sep-2011 09:46:08] <Hackman238> workers 6

[22-Sep-2011 09:46:08] <Hackman238> cachesize 100000

[22-Sep-2011 09:46:09] <Hackman238> pcachesize 2000

[22-Sep-2011 09:46:14] <Hackman238> restart each hub

[22-Sep-2011 09:47:08] <Hackman238> in $ZENHOME/etc/zenperfsnmp.conf and for each collectorname-zenperfsnmp.conf:

[22-Sep-2011 09:47:11] <Hackman238> cacheconfigs true

[22-Sep-2011 09:47:11] <Hackman238> maxqueuelen 500000

[22-Sep-2011 09:47:12] <Hackman238> eventflushchunksize 200

[22-Sep-2011 09:47:20] <Hackman238> restart each zenperfsnmp

[22-Sep-2011 09:47:28] <humanix> Hackman238, am i right?

[22-Sep-2011 09:47:58] <Hackman238> humanix: Yes and no. The devices know, but Zenoss has misconceptions

[22-Sep-2011 09:48:05] <dhopp> Hackman238:  is cacheconfigs valid for core? (I know enkrypt is enterprise..wondering if core will understand that)

[22-Sep-2011 09:48:12] <Hackman238> humanix: Zenoss will sometimes accidently choose the 32bit counters.

[22-Sep-2011 09:49:04] <Hackman238> humanix: Frankly the best way to handle is to copy the _64 template to the _32 one and put a copy of the _32 template only where it must be

[22-Sep-2011 09:49:18] <Hackman238> dhopp: I think so- I dont recall.

[22-Sep-2011 09:49:19] <humanix> but 64 can handle 32?

[22-Sep-2011 09:49:33] <dhopp> Hackman238:  looking at the code :-)

[22-Sep-2011 09:49:34] <humanix> in that case, isnt necesary to have 32 bit ones

[22-Sep-2011 09:50:10] <Hackman238> humanix: Yeah ideally you'll want to override the 32 template with a copy of the 64bit one and only put the real 32bit copy where it must be, if anywhere

[22-Sep-2011 09:50:18] <Hackman238> dhopp:

[22-Sep-2011 09:50:34] <Hackman238> dhopp: Admitadly I dont have core in front of me at this time

[22-Sep-2011 09:50:34] <dhopp>  if self.options.cacheconfigs:

[22-Sep-2011 09:50:34] <dhopp>             self.loadConfigs()

[22-Sep-2011 09:50:37] <dhopp> looks like it :-)

[22-Sep-2011 09:50:59] <Hackman238> enkrypt: Try adding those options- they will reduce zenhub load substantially

[22-Sep-2011 09:51:10] <straterra> Hackman238: Couple of issues I'm still seeing. I have that transform for unknown events (for traps), but I'm getting some SNMP traps I don't

[22-Sep-2011 09:51:17] <straterra> I don't care about ^

[22-Sep-2011 09:51:35] <Hackman238> straterra: Alrighty

[22-Sep-2011 09:51:42] <straterra> Can I put a section in the transform that drops the trap if it isn't a specific oid? I'm already matching the oid, so I guess it would just be in the else block

[22-Sep-2011 09:51:55] <enkrypt> Hackman238: ok working on it. are you sure to enable pcache? I saw somewhere that it is deprecated.

[22-Sep-2011 09:52:20] <straterra> The other issue is that when I acknowledge an issue, a page doesn't go out telling people that it's been acknowledged

[22-Sep-2011 09:52:21] <enkrypt> Hackman238: I also increased the "parallel" option in zencommand.conf and zenstatus.conf from 5 to 10, is that ok or not?

[22-Sep-2011 09:52:36] <Hackman238> enkrypt: Careful with parallel- dont have too few cores

[22-Sep-2011 09:52:50] <Hackman238> enkrypt: If must do parallel, only do on the collectors

[22-Sep-2011 09:53:16] <Hackman238> enkrypt: If we dont see zenhub aleiviated somewhat we'll look at disk IO

[22-Sep-2011 09:53:25] <enkrypt> Hackman238: ok, making the changes now

[22-Sep-2011 09:53:37] <enkrypt> do the collectors have zenhub configs in fact?

[22-Sep-2011 09:53:49] <enkrypt> I thought only the hub runs zenhub

[22-Sep-2011 09:54:44] <Hackman238> enkrypt: If you've pointed each collector to a dedicated hub, yes, if not, no

[22-Sep-2011 09:54:55] <enkrypt> because I dont have any collectername-zenhub.conf or collectorname-zenperfsnmp.conf in fact

[22-Sep-2011 09:55:04] <enkrypt> yes I've got one central hub

[22-Sep-2011 09:55:10] <enkrypt> and all collectors point to it

[22-Sep-2011 09:55:15] <enkrypt> (3 collectors atm)

[22-Sep-2011 09:55:19] <Hackman238> enkrypt: Okay skip that. We can break out to more than one hub if needed

[22-Sep-2011 09:56:13] <Hackman238> enkrypt: skip the collector-zenhub.conf's I mean. Still edit zenhub.conf

[22-Sep-2011 09:56:24] <enkrypt> by the way I'm monitoring around 2300 devices atm. all my servers (hub + collectors) have the same specs (8 physical cores, 38gb ram)

[22-Sep-2011 09:56:38] <enkrypt> most of the ram is sitting unused

[22-Sep-2011 09:56:52] <Hackman238> enkrypt: Not once we pump the caching up

[22-Sep-2011 09:57:03] <Hackman238> enkrypt: What type of disk arrays?

[22-Sep-2011 09:57:29] <enkrypt> Hackman238: 4x SAS 10k RPM 300gb disks connected to an HP raid controller with 512mb cache, battery backed

[22-Sep-2011 09:57:35] <enkrypt> RAID 1+0 that is

[22-Sep-2011 09:57:58] <Hackman238> enkrypt: Alrighty.

[22-Sep-2011 09:58:02] <enkrypt> mounted noatime and data=writeback (on the partitions where the .rrd's are)

[22-Sep-2011 09:58:24] <Hackman238> enkrypt: Are you collecting using the zenoss master or only the collectors?

[22-Sep-2011 09:58:25] <enkrypt> by the way this zenpack I'm talking about that is causing trouble, it is using zencommand

[22-Sep-2011 09:58:55] <enkrypt> it uses zencommand to run 2 scripts per device, containing some "bc" and "snmpwalk" commands

[22-Sep-2011 09:59:07] <Hackman238> enkrypt: Yep, but no matter what we do in the config for zencommand its going to suck. All we can do is minimize impact of other processes

[22-Sep-2011 09:59:14] <enkrypt> should there be some zencommand.conf tuning?

[22-Sep-2011 09:59:15] <enkrypt> ok

[22-Sep-2011 09:59:33] <enkrypt> I'm collecting only on the collectors (except for the collectors themselves, they are monitored by the master)

[22-Sep-2011 09:59:40] <Hackman238> enkrypt: You can add the same flags in the zencommand.conf's as the zenperfsnmp.conf's if you like

[22-Sep-2011 09:59:50] <Hackman238> enkrypt: Its much less dramatic an impact.

[22-Sep-2011 10:00:00] <Hackman238> enkrypt: Very good.

[22-Sep-2011 10:00:12] <enkrypt> and the parallel option for zencommand and zenstatus etc? I set it at the default of 5, or I leave at 10 ?

[22-Sep-2011 10:00:57] <Hackman238> enkrypt: I'd leave it where it was. Too few makes for long cycles. Too many causes nuts scheduler competition

[22-Sep-2011 10:02:18] <enkrypt> ok all parallel settings back to 5. changes are made on the hub, and relevant daemons restarted. I should do a collector update now for each collector?

[22-Sep-2011 10:02:48] <enkrypt> or do I need to do the config changes manually on all of the collectors?

[22-Sep-2011 10:02:49] <Hackman238> enkrypt: zenhub restarted?

[22-Sep-2011 10:03:04] <enkrypt> Hackman238: yes

[22-Sep-2011 10:03:13] <Hackman238> enkrypt: you'll want to run a collector update on each collector.

[22-Sep-2011 10:03:27] <Hackman238> enkrypt: Check afterwards that the config changes are acctually on the boxes

[22-Sep-2011 10:03:31] <enkrypt> Hackman238: I just checked the zenperfsnmp.conf etc on the collectors, and they contain only the default stuff, they don't have any cache settings? :-s

[22-Sep-2011 10:03:52] <Hackman238> enkrypt: After the update they *should* have the settings

[22-Sep-2011 10:04:08] <Hackman238> enkrypt: if they don't, just add it manually and zenoss restart on the collectors

[22-Sep-2011 10:06:50] <enkrypt> I've made some cache settings modifications previously, and now I realise the collectors never got those settings :-s

[22-Sep-2011 10:08:00] <Hackman238> enkrypt: Yeah.

[22-Sep-2011 10:12:40] <enkrypt> so that is normal procedure, that when you make config changes in /opt/zenoss/etc/*.conf you need to do these changes on all hubs+collectors manually?

[22-Sep-2011 10:13:09] <Hackman238> enkrypt: If you do a collector update and the changes dont end up on the collectors, yes

[22-Sep-2011 10:13:26] <Hackman238> enkrypt: It happens with the community distributed collector pack and the enterprise pack

[22-Sep-2011 10:13:27] <humanix> Hackman238, have you used counter64 template on snmpv1 machines?

[22-Sep-2011 10:13:40] <Hackman238> humanix: For some devices like Perles

[22-Sep-2011 10:13:59] <enkrypt> Hackman238: I'm using the enterprise distributed collector pack I assume

[22-Sep-2011 10:14:05] <Hackman238> humanix: Try to force everything in v2c mode. Last ditch is to use v1

[22-Sep-2011 10:14:14] <Hackman238> enkrypt: You run enterprise?

[22-Sep-2011 10:14:18] <Hackman238> enkrypt: or core?

[22-Sep-2011 10:16:00] <enkrypt> Hackman238: Enterprise

[22-Sep-2011 10:16:51] <Hackman238> enkrypt: Gotcha. Yeah the ent pack has the same issue sometimes.

[22-Sep-2011 10:20:43] <JohnnyNOC> Hackman238 did you say you use the WMI perf zenpack

[22-Sep-2011 10:20:44] <JohnnyNOC> ?

[22-Sep-2011 10:20:59] <Hackman238> JohnnyNOC: I don't regularly, no.

[22-Sep-2011 10:21:05] <JohnnyNOC> i've got some windows boxes for which i expect to be able tot see perf stats for physical drives, however there doesn't appear to be a template for them, and thus no graphs

[22-Sep-2011 10:21:35] <JohnnyNOC> wondering if that's the norm

[22-Sep-2011 10:23:25] <Hackman238> JohnnyNOC: Not 100% on that. Ryan or Egor would be the go to guys on that

[22-Sep-2011 10:23:52] <Hackman238> JohnnyNOC: I'm more development, network gear and super-large scale guy

[22-Sep-2011 10:24:50] <Hackman238> JohnnyNOC: I'm rpetty sure Ryan has a snmp pack to replace all WMI pack functionality plus coming

[22-Sep-2011 10:24:56] <Hackman238> JohnnyNOC: WMI just a hassle

[22-Sep-2011 10:25:15] <JohnnyNOC> yea?

[22-Sep-2011 10:25:18] <JohnnyNOC> that would be sweet

[22-Sep-2011 10:25:23] <JohnnyNOC> i agree..  wmi is a big pain in my ass

[22-Sep-2011 10:42:10] <Hackman238> JohnnyNOC:

[22-Sep-2011 10:42:15] <Hackman238> enkrypt: Luck?

[22-Sep-2011 10:46:06] <enkrypt> Hackman238: I've done an update of all collectors

[22-Sep-2011 10:46:15] <enkrypt> Hackman238: and did the .conf modifications on each of them

[22-Sep-2011 10:47:00] <Hackman238> enkrypt: Any improvement?

[22-Sep-2011 10:47:29] <enkrypt> Hackman238: cycle times seem to stay within limits (I've got command, process and perfsnmp cycle interval at 120 seconds)

[22-Sep-2011 10:47:44] <Hackman238> enkrypt: Hows the hub load look?

[22-Sep-2011 10:47:55] <enkrypt> zenhub in general seems more responsive than before

[22-Sep-2011 10:47:59] <enkrypt> I mean the GUI part

[22-Sep-2011 10:48:32] <Hackman238> enkrypt: Worst case if the hub still becomes plugged, add a new local zenhub for each collector then plug the colelctors under those hubs

[22-Sep-2011 10:49:02] <Hackman238> enkrypt: Yeah. Keep an eye on it. zenhub floods real easy- like a carborated engine with a gummy float

[22-Sep-2011 10:49:15] <enkrypt> Hackman238: well the strange thing was that even before, the system graphs for the hub were perfectly fine. almost zero CPU load, like 20gb of ram to spare, and low unix load.

[22-Sep-2011 10:49:32] <Hackman238> enkrypt: Yep thats what will happem

[22-Sep-2011 10:49:51] <Hackman238> enkrypt: Its code blocked situations that result in slow execution time rather than a load issue

[22-Sep-2011 10:50:12] <Hackman238> enkrypt: Then things pile up behind it in queue and the madness ensues

[22-Sep-2011 10:50:33] <enkrypt> Hackman238: ok so the situation is stable now, and the 1 device I have monitored using Ryan's Windows CPU+RAM zenpack shows some graphs, so that is fine. I'll try monitoring more windows machine with said zenpack now, and see if things explode

[22-Sep-2011 10:50:48] <Hackman238> enkrypt: Alrighty.

[22-Sep-2011 10:50:59] <enkrypt> it's quite shitty in fact because all of the 1200 windows machines we need to monitor only support SNMP

[22-Sep-2011 10:51:09] <Hackman238> enkrypt: Ryan will be daemonizing that pack soon to off load the work from zencommand

[22-Sep-2011 10:51:10] <enkrypt> which means, I don't have any CPU or RAM info out of the box

[22-Sep-2011 10:51:30] <Hackman238> enkrypt: That infor should be available by snmp

[22-Sep-2011 10:51:31] <enkrypt> Hackman238: yes exactly, he told me about that. that would be a god send really, because there is no way I'm activating WMI on 1200 servers

[22-Sep-2011 10:52:01] <enkrypt> Hackman238: when you add a regular Windows server (2000, 2003, 2008, whatever) it has no CPU and no MEMORY graphs. because you need to install Informant SNMP agent on those

[22-Sep-2011 10:52:13] <enkrypt> but I can't go around installing by hand Informant SNMP agent on 1200 customer servers

[22-Sep-2011 10:53:07] <Hackman238> enkrypt: I thought that too but apparently more of that data is available than thought by default

[22-Sep-2011 10:53:55] <Hackman238> enkrypt: docs/DOC-3386

[22-Sep-2011 10:54:08] <enkrypt> Hackman238: actually our current "in house developed" monitoring system does CPU + RAM just fine via SNMP. but the problem is it's kind of dynamic stuff, you can't rely on a specific OID, it depends on how many cpu's there are, and which version of windows, it's a big mess

[22-Sep-2011 10:54:31] <Hackman238> enkrypt: Agreed. I think Ryans Windows SNMP pack handles that

[22-Sep-2011 10:54:50] <enkrypt> Hackman238: yes, I'm using the Simple version of that zenpack (it gives you only TOTAL CPU instead of split up by core)

[22-Sep-2011 10:54:59] <Hackman238> enkrypt: Ah gotcha

[22-Sep-2011 10:55:06] <enkrypt> which is easier, because the Advanced pack demands that you manually put the server into a category, based on how many cpu's it has

[22-Sep-2011 10:55:20] <enkrypt> in the sense that it isn't smart enough to figure out the number of cpu's on the fly

[22-Sep-2011 10:55:27] <Hackman238> enkrypt: Yeah. I didnt know that.

[22-Sep-2011 10:55:48] <enkrypt> so any of my customers which want per-cpu graphs for their windows machines will be migrated to WMI i think

[22-Sep-2011 10:56:20] <Hackman238> enkrypt: Yeah I think WMI is needed for seperates values.

[22-Sep-2011 10:56:24] <Hackman238> enkrypt: Unfortunately

[22-Sep-2011 10:57:12] <enkrypt> but WMI brings it's own downsides, you need credentials to use it etc

[22-Sep-2011 10:57:19] <enkrypt> SNMP is easy, just set a community string and go

[22-Sep-2011 10:58:45] <Hackman238> enkrypt: Yep.

[22-Sep-2011 10:59:57] <enkrypt> Hackman238: hmmm just noticed zenvmwaremodeler had trouble connecting to zenhub, 1 timeout, and then he did succeed

[22-Sep-2011 11:00:25] <hikenboot> ok I figured out what the problem was I now have net-snmp-agent working its actually nexentastor that I am monitoring which is loosely based on solaris kernel with debian userland...anyways it has a command for it setup network service snmp-agent configure

[22-Sep-2011 11:00:57] <enkrypt> Hackman238: another question, when using Enterprise and distributed collectors, where does zenrender need to be running? because by default on the hub it doesn't start, and afterwards I get warnings in the dashboard saying that the zenrender heartbeat is missing

[22-Sep-2011 11:01:43] <cluther> enkrypt: zenrender should only run on remote collectors.

[22-Sep-2011 11:02:05] <cluther> enkrypt: The render URL for local collectors should be /zport/RenderServer which uses Zope to render.

[22-Sep-2011 11:02:26] <enkrypt> cluther: ok gotcha

[22-Sep-2011 11:02:34] <enkrypt> there we go again, my zenhub is sinking

[22-Sep-2011 11:03:02] <enkrypt> the GUI itself remains responsive, but I can't show anymore graphs (graphs residing on remote collector)

[22-Sep-2011 11:04:00] <enkrypt> zenhub is still running, but the log file doesn't grow anymore (even though I activated debug)

[22-Sep-2011 11:04:35] <cluther> enkrypt: Did you just make any changes to zProperties or monitoring templates? What version of Zenoss are you on?

[22-Sep-2011 11:05:17] <enkrypt> cluther: no changes made. I'm on Enterprise 3.2.0 + the 3 patches listed in the topic

[22-Sep-2011 11:05:34] <enkrypt> cluther: Hackman238 helped me tweaking some settings in zenperfsnmp.conf to cache configs etc

[22-Sep-2011 11:05:57] <enkrypt> cluther: I'm trying to use the Windows CPU/RAM zenpack created by Ryan Matte, to provide CPU and RAM graphs for windows hosts

[22-Sep-2011 11:06:03] <enkrypt> cluther: which is based on zencommand

[22-Sep-2011 11:06:17] <enkrypt> cluther: but as soon as I monitoring more than 2 servers with that zenpack, my zenhub goes nuts

[22-Sep-2011 11:07:53] <enkrypt> I've got quite powerful hardware, and almost no cpu or unix load, and nearly 30 gigabyte of available ram, but as soon as I activate this zenpack my zenhub simply goes in the corner and shoots itself

[22-Sep-2011 11:08:35] <enkrypt> which is a bit problematic because this means that I have to tell the owners of these 1200 windows servers that they can no longer have CPU or Memory graphs with our state of the art monitoring software :'(

[22-Sep-2011 11:09:40] <cluther> How many workers is your hub configured for?

[22-Sep-2011 11:10:20] <enkrypt> cluther: workers = 6, cachesize 100000, pcachesize 2000

[22-Sep-2011 11:10:29] <enkrypt> 8 physical cpu cores and 38gb ram

[22-Sep-2011 11:10:52] <cluther> Which ZenPack is that specifically? I'd like to take look at the thresholds defined in it.

[22-Sep-2011 11:11:17] <enkrypt> docs/DOC-3570

[22-Sep-2011 11:11:36] <enkrypt> the problematic part is that it uses zencommand for running 2 scripts per device it monitors I think

[22-Sep-2011 11:11:51] <Hackman238> enkrypt: zenreder on each collector

[22-Sep-2011 11:11:53] <enkrypt> maybe I should do some tweaking with respect to parallel = ? or zencommand.conf, or elsewhere, I'm not sure what to do

[22-Sep-2011 11:12:07] <enkrypt> Hackman238: but not on the hub right? ok thx.

[22-Sep-2011 11:12:38] <Hackman238> enkrypt: on the daemon perf page are the event queues high?

[22-Sep-2011 11:12:57] <enkrypt> Hackman238: not really

[22-Sep-2011 11:13:01] <Hackman238> Hey cluther, how goes it?

[22-Sep-2011 11:13:18] <cluther> enkrypt: I think I see the problem. The "Memory" threshold's maximum expression is "here.getRRDValue('MemoryTotal') * 0.9"

[22-Sep-2011 11:13:21] <enkrypt> highest is a few hundred for zensnmp

[22-Sep-2011 11:13:31] <Hackman238> enkrypt: I'd break up your zenhub in to 3, one for everythig local, one for each collector

[22-Sep-2011 11:13:33] <cluther> Hackman238: Great. Thanks for asking. You?

[22-Sep-2011 11:13:52] <Hackman238> cluther: Very well

[22-Sep-2011 11:13:58] <cluther> enkrypt: zenhub is deadlocking trying to make an XML-RPC call to itself.

[22-Sep-2011 11:14:19] <cluther> enkrypt: What is the render URL for one of your remote collectors? Can you paste it with hostname/IP obscured if needed?

[22-Sep-2011 11:14:21] <Hackman238> cluther: Blocking code then a pile up?

[22-Sep-2011 11:14:50] <cluther> Hackman238: Imagine a synchronous web server trying to retrieve a page from itself to serve a page request.

[22-Sep-2011 11:14:52] <enkrypt> looking it up

[22-Sep-2011 11:15:09] <enkrypt> http://zenoss1.cegeka.be:8090/collectorHA01

[22-Sep-2011 11:15:23] <enkrypt> zenoss1.cegeka.be being the HUB

[22-Sep-2011 11:15:33] <cluther> enkrypt: Do your users (their browsers) have network access through to port 8091/tcp on your remote collectors? Or do you have network security constraints?

[22-Sep-2011 11:15:46] <Hackman238> enkrypt: You'll want to change that reder url

[22-Sep-2011 11:16:00] <enkrypt> cluther: I can open up anything that is needed

[22-Sep-2011 11:16:15] <Hackman238> enkrypt: http://yourremotecollector:8091/

[22-Sep-2011 11:16:15] <cluther> enkrypt: Change all remote collectors' render URLs to http://collector.hostname:8091/

[22-Sep-2011 11:16:21] <enkrypt> is the render URL wrong? I don't remember changing it

[22-Sep-2011 11:16:21] <Hackman238> enkrypt: For each of the collector render url

[22-Sep-2011 11:16:36] <enkrypt> I do have to say that I have had since the beginning very weird issues with showing graphs

[22-Sep-2011 11:16:45] <cluther> enkrypt: It's not wrong. It's just another way of doing it that works better in environments with more network security in place.

[22-Sep-2011 11:16:55] <cluther> enkrypt: This is going to solve all of those problems!

[22-Sep-2011 11:17:13] <enkrypt> sometimes they simply are not shown, giving me the famous error message telling me to wait 5 minutes or to look in $ZENHOME/log/event.log. and then i refresh, and the graph is there. and then i refresh again, and the error is back

[22-Sep-2011 11:17:25] <enkrypt> very spooky

[22-Sep-2011 11:17:45] <enkrypt> ok I'll try changing them everywhere

[22-Sep-2011 11:17:53] <enkrypt> do I need to do a collector update afterwards?

[22-Sep-2011 11:18:06] <Hackman238> No

[22-Sep-2011 11:18:48] <Hackman238> cluther: Gotcha

[22-Sep-2011 11:19:17] <enkrypt> oh man if this would solve my problems I would be a happy camper

[22-Sep-2011 11:19:34] <enkrypt> I need to go live with this project by the end of the year and I've still got a tonne of things to work out

[22-Sep-2011 11:21:34] <enkrypt> ok it's blocked by the firewall, will need to adjust that, brb

[22-Sep-2011 11:22:01] <cluther> enkrypt: The Zenoss master server and all users of the web interface will need to be able to hit all collectors on 8091/tcp.

[22-Sep-2011 11:22:15] <rmatte> enkrypt: was it you who had the problem monitoring OS processes with 3.2?

[22-Sep-2011 11:22:20] <enkrypt> after this change do my users still need to be able to go port 8090 anywhere?

[22-Sep-2011 11:22:28] <cluther> enkrypt: No.

[22-Sep-2011 11:22:29] <enkrypt> rmatte: yes that's me

[22-Sep-2011 11:22:41] <rmatte> enkrypt: http://dev.zenoss.org/trac/ticket/7870#comment:5 They can't reproduce the issue.

[22-Sep-2011 11:22:57] <enkrypt> rmatte: I'm still struggling to monitoring windows machines with your zenpack, cluther and Hackman238 are helping me out

[22-Sep-2011 11:23:14] <rmatte> enkrypt: perhaps you could post the exact options that each script is running with?

[22-Sep-2011 11:23:27] <rmatte> It could be something very specific to the way the options are formatted or something

[22-Sep-2011 11:24:00] <rmatte> enkrypt: My ZenPack uses a command datasource and you were looking to monitor like 3000 hosts with it, it's not going to work unless you have some serious hardware

[22-Sep-2011 11:24:10] <rmatte> though you said you'd try it against 20 devices first

[22-Sep-2011 11:24:20] <enkrypt> and the graphs work!

[22-Sep-2011 11:24:21] <rmatte> which should be absolutely braindead simple to do

[22-Sep-2011 11:24:54] <enkrypt> rmatte: I got it working with 1 host your zenpack. I said ok cool. then I tried with 20 hosts, and my zenhub went into the corner and cried like a baby

[22-Sep-2011 11:25:12] <rmatte> sounds like your hardware isn't up to the task or something

[22-Sep-2011 11:25:18] <rmatte> or you need to tune your Zenoss

[22-Sep-2011 11:25:41] <rmatte> I've got that script working against over 100 servers on some of my Zenoss boxes without issue

[22-Sep-2011 11:25:43] <enkrypt> rmatte: according to cluther and Hackman238 it might have had something to do with my render URL

[22-Sep-2011 11:25:51] <rmatte> ah

[22-Sep-2011 11:25:56] <rmatte> that could be

[22-Sep-2011 11:26:01] <enkrypt> cluther: didn't you say something about the Threshold that might be a problem?

[22-Sep-2011 11:26:20] <cluther> It's the Memory threshold with the maximum expression of "here.getRRDValue('MemoryTotal') * 0.9"

[22-Sep-2011 11:26:30] <rmatte> the threshold uses getRRDValue, so that needs to be able to work for it to function correctly

[22-Sep-2011 11:26:42] <enkrypt> aha!

[22-Sep-2011 11:26:51] <rmatte> if you can't return a getRRDValue then it won't work

[22-Sep-2011 11:26:57] <cluther> With the zenhub:8090-style render URL this will cause zenhub to make a request to itself to fulfill the getRRDValue call while building a configuration for zencommand. This deadlocks zenhub.

[22-Sep-2011 11:27:01] <rmatte> but if you can't return a getRRDValue then you have bigger problems

[22-Sep-2011 11:27:18] <rmatte> cluther: ah

[22-Sep-2011 11:27:31] <rmatte> so what's the style of URL required?

[22-Sep-2011 11:27:43] <rmatte> 8091?

[22-Sep-2011 11:28:12] <cluther> It's dead(lock)y to use getRRDValue in threshold expressions or event transforms in distributed Zenoss setups unless you use the collector:8091/-style render URLs.

[22-Sep-2011 11:28:19] <cluther> rmatte: Yeah.

[22-Sep-2011 11:28:25] <rmatte> ah ok

[22-Sep-2011 11:28:51] <rmatte> I don't even get why there are 2 different types of render URL

[22-Sep-2011 11:28:59] <rmatte> 8091 seems to be the less problematic of the two

[22-Sep-2011 11:29:05] <rmatte> shouldn't that basically be the standard?

[22-Sep-2011 11:29:07] <enkrypt> cluther: and does this explain also why every so often I used to get error messages telling me to wait 5 minutes and go look in event.log, instead of seeing graphs?

[22-Sep-2011 11:29:21] <cluther> 8091 is problematic if your users don't have network access to the collectors. This is pretty common in my experience.

[22-Sep-2011 11:29:41] <rmatte> ah

[22-Sep-2011 11:29:50] <cluther> 8090 lets you tunnel all requests through your matter which, by definition, has to be network-accessible to users.

[22-Sep-2011 11:29:54] <rmatte> some 8091 is direct access and 8090 is relayed?

[22-Sep-2011 11:29:57] <rmatte> so*

[22-Sep-2011 11:29:59] <cluther> yup

[22-Sep-2011 11:30:05] <rmatte> I see

[22-Sep-2011 11:30:46] <rmatte> interesting, good to know

[22-Sep-2011 11:30:50] <cluther> I agree with you about 8091 though. It's so much better if you can use it. Rendering is much faster and you don't put extra load on poor zenhub.

[22-Sep-2011 11:31:07] <rmatte> for sure

[22-Sep-2011 11:31:32] <Hackman238> cluther: Very interesting.

[22-Sep-2011 11:31:37] <Hackman238> rmatte: how goes it?

[22-Sep-2011 11:32:06] <rmatte> Hackman238: alright, been sick the past few days

[22-Sep-2011 11:32:18] <rmatte> took Tuesday off and worked sparingly yesterday

[22-Sep-2011 11:32:35] <rmatte> feeling better today

[22-Sep-2011 11:32:45] <enkrypt> ok I'm trying to add around 20 devices to the zenpack now, lets see if it chokes

[22-Sep-2011 11:33:29] <rmatte> enkrypt: for your OS process monitoring issue, was that when you tested against devices on a collector, or were they directly on the hub?

[22-Sep-2011 11:33:49] <rmatte> Ian needs more details since his testing so far hasn't revealed any issue.

[22-Sep-2011 11:34:04] <Hackman238> rmatte: Glad youre feeling better

[22-Sep-2011 11:34:09] <rmatte> Hackman238: thanks

[22-Sep-2011 11:34:42] <rmatte> My main assignment for today is writing scripts to dump and compute Zenoss data for a yearly utilization report that we provide to one of our clients

[22-Sep-2011 11:35:06] <rmatte> probably going to take a good portion of the day to do

[22-Sep-2011 11:35:19] <rmatte> since I have to modify a bunch of my reports scripts to do it

[22-Sep-2011 11:35:29] <rmatte> they want the utilization info per device rather than per component

[22-Sep-2011 11:35:38] <rmatte> so total interface errors per device or whatever

[22-Sep-2011 11:35:45] <Hackman238> rmatte: I wrote one: message/60932#60932

[22-Sep-2011 11:36:00] <Hackman238> rmatte: Jumping off point if you like

[22-Sep-2011 11:36:25] <enkrypt> rmatte: the devices for that bug were in both cases in fact. some of them monitored through the hub, some of them through the collector. the OSPRocess monitoring was set up by David Carmean (from Zenoss), and it worked fine in 3.1.0, but broke in 3.2.0

[22-Sep-2011 11:36:31] <rmatte> did you use my scripts as a base for that?

[22-Sep-2011 11:36:36] <rmatte> the option lines look familiar

[22-Sep-2011 11:36:37] <rmatte> lol

[22-Sep-2011 11:37:50] <rmatte> enkrypt: ok, well I'd suggest that you add as much data as humanly possible to the ticket so that ian has something to work with (including the options that each process is running with): http://dev.zenoss.org/trac/ticket/7870#comment:5

[22-Sep-2011 11:39:11] <Hackman238> rmatte: No idea. I wrote is so long ago. I wouldn't be suprised

[22-Sep-2011 11:39:41] <rmatte> I think you based it off these: docs/DOC-4664

[22-Sep-2011 11:40:14] <rmatte> but anyways, thanks, this saves me some time

[22-Sep-2011 11:40:34] <enkrypt> rmatte: ok I'll update the ticket

[22-Sep-2011 11:40:42] <rmatte> enkrypt: thank you

[22-Sep-2011 11:41:00] <Hackman238> rmatte: Those are too new, but I think I used a script of yours I found in a post.

[22-Sep-2011 11:41:18] <enkrypt> by the way, I'm really not so happy with what David Carmean set up on this end. yesterday I had a debugging session with Eric from support, and he found out that my entire vmware monitoring was messed up, because David removed the -1 production state from my zenoss installation

[22-Sep-2011 11:41:18] <Hackman238> rmatte: always recycle work where possible LOL

[22-Sep-2011 11:41:39] <enkrypt> and now apparently I found out that I'm using a sub-optimal render URl which causes many things to go wrong

[22-Sep-2011 11:41:46] <Hackman238> enkrypt: Whos Davind Carmean?

[22-Sep-2011 11:41:51] <rmatte> Hackman238: agreed lol

[22-Sep-2011 11:42:32] <rmatte> Hackman238: ah, yeh, they've been around a lot longer than that stand-alone reports page.  I just added them there later on and Matt made them a tip of the month

[22-Sep-2011 11:42:44] <Hackman238> enkrypt: If he works for Zenoss support, email experiences to rusty@zenoss.com.

[22-Sep-2011 11:42:45] <rmatte> but they were around way before that and I'm sure I posted them somewhere at some point

[22-Sep-2011 11:43:01] <Hackman238> rmatte: Yep

[22-Sep-2011 11:43:34] <enkrypt> Hackman238: he's a new guy working for Zenoss, he is the one that helped us for a few weeks setting up the project. cost us around 30.000 USD, but in the end I am discovering that many things don't work well, and there is many work left undone

[22-Sep-2011 11:43:47] <enkrypt> a lot of work *

[22-Sep-2011 11:44:09] <Hackman238> enkrypt: Hit up rusty.

[22-Sep-2011 11:44:31] <rmatte> yeh, rusty will sort you out

[22-Sep-2011 11:44:41] <Hackman238> enkrypt: Generally speaking Eric and Zack are the best support people.

[22-Sep-2011 11:45:04] <rmatte> well, technically Chet is, but he's not so much support as PS

[22-Sep-2011 11:45:09] <Hackman238> enkrypt: Developer wise, Chet and the others are all dark mages of Python and Zope

[22-Sep-2011 11:45:09] <rmatte>

[22-Sep-2011 11:45:25] <Hackman238> enkrypt: Any of them know everything inside-out

[22-Sep-2011 11:45:30] <Hackman238> rmatte: LOL yep

[22-Sep-2011 11:46:09] <enkrypt> yes I did a session with Eric 2 days ago, he helped me out

[22-Sep-2011 11:46:30] <Hackman238> enkrypt: Glad to hear

[22-Sep-2011 11:46:49] <enkrypt> he was amazed at the errors that were to be found in my system, such as the deletion of the -1 production state

[22-Sep-2011 11:46:56] <enkrypt> he was going to talk to someone internally about that

[22-Sep-2011 11:47:36] <rmatte> lol

[22-Sep-2011 11:47:59] <rmatte> no one should touch prod states unless they know what they're doing

[22-Sep-2011 11:48:10] <Hackman238> enkrypt: Madness!

[22-Sep-2011 11:48:12] <Hackman238> rmatte: Agreed.

[22-Sep-2011 11:49:06] <cluther> General rule of thumb: if you don't understand something, don't delete it..

[22-Sep-2011 11:49:34] <cluther> Brings back memories of Windows tech support and people deleting .dll files.

[22-Sep-2011 11:49:46] <rmatte> haha, so true

[22-Sep-2011 11:50:16] <Hackman238> Very very true

[22-Sep-2011 11:50:46] <Hackman238> *recalls back in the day calling people and telling them to deltree c:\windows /y because they had a virus*

[22-Sep-2011 11:50:47] <enkrypt> ok I'm hitting problems again. I try to look at the Performance graphs for my collector.

[22-Sep-2011 11:51:02] <enkrypt> some of the graphs work, other show the error message saying me to wait 5 minutes and look in event.log

[22-Sep-2011 11:51:15] <enkrypt> I hit refresh, now those graphs work, but OTHERS show the error message

[22-Sep-2011 11:51:30] <enkrypt> very spooky, at just random times it fails to show certain graphs

[22-Sep-2011 11:51:35] <Hackman238> enkrypt: Thats bizarre. And the render url's are updated?

[22-Sep-2011 11:51:43] <enkrypt> yes, render URL updated everywhere

[22-Sep-2011 11:52:00] <enkrypt> I checked with my browser debugger (chrome) he goes to the correct URL now, port 8091

[22-Sep-2011 11:52:13] <Hackman238> enkrypt: No round robbin A records, right?

[22-Sep-2011 11:52:24] <enkrypt> Hackman238: right

[22-Sep-2011 11:53:37] <cluther> enkrypt: That is strange.. sounds like caching somewhere. Blow away your browser cache and restart the *_zenrender daemons if that doesn't do it.

[22-Sep-2011 11:53:38] <dhopp> is there a way to have zenoss send all network requests from a specific IP?

[22-Sep-2011 11:55:32] <Hackman238> dhopp: What do you mean?

[22-Sep-2011 11:56:08] <rmatte> Hackman238: if he has 2 interfaces with even routing between them (load balancing) on a server he wants to use one specific interface for sending Zenoss traffic from

[22-Sep-2011 11:56:12] <rmatte> I believe is what he means

[22-Sep-2011 11:56:12] <nyeates> rmatte you around?

[22-Sep-2011 11:56:13] <nyeates> ok u are

[22-Sep-2011 11:56:22] <dhopp> Hackman238:  for example..when zenoss goes to poll a server with snmp I would like the request to come from a specific IP on the zenoss box rather than from the eth0 IP

[22-Sep-2011 11:56:26] <rmatte> and I believe the answer is to change the "monitor host" option for each daemon

[22-Sep-2011 11:56:40] <nyeates> on your bug 7870 with process monitoring regex's..... i havent had any time to try this myself..... are you still seeing it on your box?

[22-Sep-2011 11:56:49] <dhopp> rmatte:  sort of…we have eth0 and eth0:svc..I want all requests to come from eth0:svc

[22-Sep-2011 11:57:01] <cluther> dhopp: Is this for an HA setup? If so, and you're using the Linux-HA project you probably want to look at the IPsrcaddr resource.

[22-Sep-2011 11:57:05] <rmatte> dhopp: change the monitor host option for each daemon

[22-Sep-2011 11:57:06] <Hackman238> dhopp: use bonding rather than two ints and a CSS

[22-Sep-2011 11:57:06] <cluther> dhopp: It'll do it for you on failover.

[22-Sep-2011 11:57:42] <dhopp> cluther:  no..it's more for traversing NATs and firewall rules

[22-Sep-2011 11:57:51] <rmatte> nyeates: that's not actually my bug, that's enkrypt's bug

[22-Sep-2011 11:58:00] <enkrypt> cluther: it keeps doing it now and then

[22-Sep-2011 11:58:06] <rmatte> nyeates: I'm just helping to keep track of it, I requested that he update the ticket with more detailed information

[22-Sep-2011 11:58:20] <enkrypt> cluther: I ask some graphs, and some of them are missing, giving me error message. i refresh, and all is ok. I refresh again, and some missing again

[22-Sep-2011 11:58:26] <Hackman238> nyeates: Hey, how goes it?

[22-Sep-2011 11:58:29] <cluther> dhopp: http://www.readmespot.com/question/f/127636/force-local-ip-traffic-to-an-external-interface

[22-Sep-2011 11:58:31] <enkrypt> cleaned cache completely, and restarted all zenrender daemons

[22-Sep-2011 11:59:03] <nyeates> enkrypt: add yourself as a cc email contact on bug http://dev.zenoss.com/trac/ticket/7870

[22-Sep-2011 11:59:17] <nyeates> enkrypt: you can login as zenoss/zenoss to do this

[22-Sep-2011 11:59:44] <enkrypt> nyeates: did it

[22-Sep-2011 11:59:55] <enkrypt> thx

[22-Sep-2011 12:00:01] [disconnected at Thu Sep 22 12:00:01 2011]

[22-Sep-2011 12:00:02] [connected at Thu Sep 22 12:00:02 2011]

[22-Sep-2011 12:00:11] [zenoss-logger (logger bot) has joined #zenoss]

[22-Sep-2011 12:00:14] <enkrypt> There was a problem rendering this graph. Either the file does not exist or an error has occurred. Initial graph creation can take up to 5 minutes. If the graph still does not appear, look in the Zope log file $ZENHOME/log/event.log for errors.

[22-Sep-2011 12:00:25] <enkrypt> ^^^^ this is the error I get to see at random times, when watching graphs.

[22-Sep-2011 12:00:33] <dhopp> cluther:  that will just route the packet out that interface..it won't necessarily use that interface IP as the source IP though?

[22-Sep-2011 12:00:52] <dhopp> oh..I didn't read down through the iptables stuff

[22-Sep-2011 12:01:05] <dhopp> that could get really messy and make it troublesome to debug

[22-Sep-2011 12:03:10] <cluther> dhopp: I haven't tested the "ip route change to" method myself, but I would think sending a packet out a specific interface would set its source IP that that interface's IP. I could be wrong about that.

[22-Sep-2011 12:03:49] <Jane_Curry> Hi cluther

[22-Sep-2011 12:04:08] <cluther> dhopp: In fact I would bet that the "ip route change to" method changes the source IP, since I got it from Linux-HA's IPsrcaddr resource script and this is the purpose of the script.

[22-Sep-2011 12:04:10] <Jane_Curry> Any chance you could take a quick look at thread/16821 and comment??

[22-Sep-2011 12:04:28] <cluther> dhopp: I know that the iptables POSTROUTING method works.

[22-Sep-2011 12:04:33] <cluther> Jane_Curry: Looking..

[22-Sep-2011 12:05:38] <Hackman238> Jane_Curry: the old PT overrides only work for legacy pages, not the js pages

[22-Sep-2011 12:05:56] <Hackman238> Jane_Curry: I'm not sure about the zcml override. Im sure chet knows

[22-Sep-2011 12:06:56] <cluther> Shane's right about it being defined in ZCML now. Specifically if you search for editActionRule in Products/ZenUI3/browser/backcompat.zcml.

[22-Sep-2011 12:07:21] <Jane_Curry> But the page I want to override (modifying an alerting rule) I think is fundamentally on old-style page

[22-Sep-2011 12:08:15] <cluther> It is an old-style page, but we're still registering it in the new more specific way instead of relying on the old skin layers.

[22-Sep-2011 12:08:40] <Jane_Curry> even though it is in backcompat.zcml

[22-Sep-2011 12:08:44] <cluther> I'll post the ZCML you need. I'm just testing it first.

[22-Sep-2011 12:08:59] <Jane_Curry> Great

[22-Sep-2011 12:10:03] <Hackman238> cluther: Thanks

[22-Sep-2011 12:20:31] <Jane_Curry> cluther: If you happened to have a code snippet that answered message/61431#61431 that would be awesome too

[22-Sep-2011 12:23:21] <enkrypt> rmatte: seems like the windows cpu zenpack is giving much less problems after I changed the Render URL as cluther / Hackman238 suggested

[22-Sep-2011 12:26:07] <cluther> Jane_Curry: I posted a response on that first thread.

[22-Sep-2011 12:28:22] <enkrypt> thanks guys for the help, I'm off for today (it's 6:30pm over here). skipping town tomorrow, spending the weekend in Paris see you next week, enjoy the weekend!

[22-Sep-2011 12:28:34] <cluther> Jane_Curry: Doh! The overriding security question. I forgot about that thread. I don't know how to do it off-hand and was half through researching it.

[22-Sep-2011 12:30:04] <JohnnyNOC> are os processes broken in Zenoss 3.1.0?

[22-Sep-2011 12:30:20] <JohnnyNOC> my collector plugins look to be collecting the data however no 'os process' component is ever created

[22-Sep-2011 12:31:01] <JohnnyNOC> i see this on both windows and linux machines :/

[22-Sep-2011 12:31:23] <JohnnyNOC> with windows i use community.wmi.process and when i run the modeler with just that plugin i see all the processes

[22-Sep-2011 12:31:52] <JohnnyNOC> similar for linux using zenoss.snmp.HRSWRunMap..  i can run that with zenmodeler and see the processes but i don't get 'os processes' under components

[22-Sep-2011 12:32:53] <rmatte> JohnnyNOC: as far as I recall there was a patch for that...

[22-Sep-2011 12:33:04] <rmatte> let me dig it up...

[22-Sep-2011 12:33:04] <Jane_Curry> cluther: thanks for the post - I'll have a go at that - much appreciated

[22-Sep-2011 12:33:10] <JohnnyNOC> rmatte serious?  dang

[22-Sep-2011 12:33:18] <Jane_Curry> cluther: Great to see you here 

[22-Sep-2011 12:34:07] <rmatte> JohnnyNOC: ah, my mistake, it was for IP Service Monitor, not OS process

[22-Sep-2011 12:34:27] <JohnnyNOC> ah

[22-Sep-2011 12:35:06] <JohnnyNOC> well, i must be doing something wrong

[22-Sep-2011 12:35:25] <JohnnyNOC> is there something else i should be taking into considering besides the collector plugins?

[22-Sep-2011 12:35:28] <rmatte> JohnnyNOC: you have actually defined the processes in the processes section right?

[22-Sep-2011 12:35:34] <JohnnyNOC> no

[22-Sep-2011 12:35:38] <JohnnyNOC> i have not

[22-Sep-2011 12:35:51] <rmatte> That's required, it doesn't just pick up any process, you have to tell it which ones you want it to detect and monitor

[22-Sep-2011 12:35:54] <JohnnyNOC> my hope was it would be like the services....

[22-Sep-2011 12:35:59] <rmatte> no, it's not

[22-Sep-2011 12:36:01] <JohnnyNOC> ok, well i guess that explains it

[22-Sep-2011 12:36:02] <JohnnyNOC>

[22-Sep-2011 12:36:06] <rmatte> that would be ridiculous

[22-Sep-2011 12:36:12] <rmatte> there are hundreds of processes

[22-Sep-2011 12:36:15] <JohnnyNOC> *nod* i could see that

[22-Sep-2011 12:36:27] <JohnnyNOC> sure but i have some dvices with 100+ windows services it finds

[22-Sep-2011 12:36:39] <JohnnyNOC> so yes, i was incorrect, but i assumed it would be similar with the os processes

[22-Sep-2011 12:36:52] <rmatte> Infrastructure -> Processes

[22-Sep-2011 12:36:55] <rmatte> you need to add them in there

[22-Sep-2011 12:36:57] <JohnnyNOC> *nod*

[22-Sep-2011 12:37:32] <JohnnyNOC> process added..  remodeling

[22-Sep-2011 12:37:37] <JohnnyNOC> let's see what happens

[22-Sep-2011 12:37:37] <JohnnyNOC>

[22-Sep-2011 12:38:00] <JohnnyNOC> yep

[22-Sep-2011 12:38:04] <rmatte> also keep in mind that you may want to enable the option to ignore parameters for the processes

[22-Sep-2011 12:38:26] <rmatte> for instance, of there are 10 processes running for the same thing, apache for instance

[22-Sep-2011 12:38:28] <JohnnyNOC> is that in config properties?

[22-Sep-2011 12:38:37] <rmatte> you just want to make sure there's at least 1 process, not that each individual is running

[22-Sep-2011 12:38:44] <rmatte> so you'd enable ignore parameters

[22-Sep-2011 12:38:55] <rmatte> it's an option right on the process definition itself

[22-Sep-2011 12:39:06] <JohnnyNOC> ah

[22-Sep-2011 12:39:08] <JohnnyNOC> gotcha i see it

[22-Sep-2011 12:39:29] <rmatte> I use that option a lot

[22-Sep-2011 12:42:00] <JohnnyNOC> as always, thanks for the help

[22-Sep-2011 12:42:00] <JohnnyNOC>

[22-Sep-2011 12:42:06] <rmatte> no problem

[22-Sep-2011 12:43:58] <Hackman238> ls

[22-Sep-2011 12:46:28] <dhopp> boot  etc  lib

[22-Sep-2011 12:46:40] <dhopp> Hackman238:  you need a faster ps..your ls took forever to return :-P

[22-Sep-2011 12:46:44] <dhopp> *pc

[22-Sep-2011 12:47:06] <rmatte> you're supposed to do...

[22-Sep-2011 12:47:07] <rmatte> .

[22-Sep-2011 12:47:08] <rmatte> ..

[22-Sep-2011 12:47:15] <rmatte> /bin

[22-Sep-2011 12:47:17] <rmatte> /usr

[22-Sep-2011 12:47:18] <rmatte> lol

[22-Sep-2011 12:47:42] <Hackman238> LOL

[22-Sep-2011 12:47:46] <nexex> he didn't have -a

[22-Sep-2011 12:48:00] <rmatte> true

[22-Sep-2011 12:48:18] <dhopp> rmatte:  he had it aliased on my server

[22-Sep-2011 12:48:27] <dhopp> err..nexex I mean

[22-Sep-2011 12:48:38] <rmatte> in case anyone doesn't know, -1 is the most useful ls option for scripting

[22-Sep-2011 12:48:41] <rmatte> ls -1

[22-Sep-2011 12:48:43] <rmatte> ls -a1

[22-Sep-2011 12:48:52] <dhopp> rmatte: ls −1 -U :-

[22-Sep-2011 12:48:53] <rmatte> that's the number one

[22-Sep-2011 12:48:54] <dhopp> :-)

[22-Sep-2011 12:49:11] <rmatte> lol

[22-Sep-2011 12:49:23] <dhopp> I had a directory with 2 million files in it…ls −1 -U actually made it so could figure out wtf was happening

[22-Sep-2011 12:49:41] <rmatte> nice

[22-Sep-2011 12:50:37] <rmatte> eugh, now I need to decypher this date line that I wrote ages ago...

[22-Sep-2011 12:50:46] <rmatte> date -d "-1 month -$((10#$(date +%d)-1)) days -$((10#$(date +%H)-1)) hours -$((10#$(date +%M))) minutes -$((10#$(date +%S))) seconds" +%s

[22-Sep-2011 12:51:44] <rmatte> it gets the date from 1 month ago in epoch time

[22-Sep-2011 12:51:54] <rmatte> but I forget why I added all that extra stuff to it

[22-Sep-2011 12:52:58] <rmatte> oh right

[22-Sep-2011 12:53:16] <rmatte> -1 month and then however many days, hours, minutes, and seconds currently exist past that

[22-Sep-2011 12:55:31] <rmatte> and the 10# was something to do with the way it was actually handling the values

[22-Sep-2011 13:03:18] <cluther> Jane_Curry: Thank you for making me learn something new and awesome today!

[22-Sep-2011 13:03:58] <Hackman238> cluther: Defiantely something we all needed to learn

[22-Sep-2011 13:04:24] <cluther> Jane_Curry: I've been banging my head on the wall over the overriding security question. It turns out there's a very simple answer that will make your life a lot easier.

[22-Sep-2011 13:06:16] <cluther> I just put the info into the forum thread. message/61431

[22-Sep-2011 13:07:03] <Hackman238> cluther: Wow, thats a really simple solution. Nice!

[22-Sep-2011 13:07:57] <cluther> ZCA can be your friend. Zope Component Architecture that is..

[22-Sep-2011 13:09:28] <Hackman238> cluther: I didnt even think to look in my zope programming pdf. Damn

[22-Sep-2011 13:09:34] <Hackman238> *face palm*

[22-Sep-2011 13:09:43] <Hackman238> cluther: Thanks again

[22-Sep-2011 13:09:54] <rmatte> Hackman238: if you still have that script laying around can you pastebin it?  Jive jumbles it all to shit when I copy/paste it.

[22-Sep-2011 13:10:08] <Hackman238> rmatte: Sure

[22-Sep-2011 13:10:11] <rmatte> thanks

[22-Sep-2011 13:11:33] <Hackman238> rmatte: http://pastebin.com/zgriGZs7

[22-Sep-2011 13:12:03] <rmatte> thanks

[22-Sep-2011 13:12:11] <Hackman238> rmatte: np

[22-Sep-2011 13:13:00] <Jane_Curry> cluther: wow! Thanks. Will try em!

[22-Sep-2011 13:13:07] <jb> hey everyone.

[22-Sep-2011 13:14:50] <Hackman238> jb: Hey

[22-Sep-2011 13:15:08] <jb> hopefully the disks for my new 4.0.2 box came in today

[22-Sep-2011 13:15:35] <jb> i'm hoping the 4.0.2 -> 4.2 upgrade will be smooth.

[22-Sep-2011 13:15:43] <jb> support hasn't answered yet.

[22-Sep-2011 13:16:01] <Hackman238> jb: you and me both LOL

[22-Sep-2011 13:49:47] <humanix> rmatte, hi man

[22-Sep-2011 13:49:58] <rmatte> yo

[22-Sep-2011 13:50:04] <humanix> howre u?

[22-Sep-2011 13:50:13] <rmatte> good

[22-Sep-2011 13:50:43] kocolosk|away is now known as kocolosk

[22-Sep-2011 13:50:59] <humanix> man i got a doubt about zenmodel

[22-Sep-2011 13:51:13] <humanix> does zenmodel run automatically on zenoss 2.x? or you have to manually model

[22-Sep-2011 13:51:16] <humanix> always?

[22-Sep-2011 13:51:26] <rmatte> it runs automatically by default

[22-Sep-2011 13:51:34] <rmatte> every 720 minutes I believe is the default setting

[22-Sep-2011 13:51:38] <rmatte> but it can be changed

[22-Sep-2011 13:51:48] <rmatte> I personally disable the automatic remodelling

[22-Sep-2011 13:52:12] <humanix> why?

[22-Sep-2011 13:52:32] <rmatte> because we monitor across several VPN links, many of which go to europe and are slow/low bandwidth

[22-Sep-2011 13:52:39] <rmatte> so I only model when necessary

[22-Sep-2011 13:52:58] <rmatte> also, I don't want new interfaces and stuff just magically picked up in the middle of the night causing phone calls

[22-Sep-2011 13:53:11] <rmatte> when customers make changes they let us know and we remodel by hand, that's our process

[22-Sep-2011 13:53:12] <humanix> where are you currently?

[22-Sep-2011 13:53:17] <Hackman238> cluther: You have 5?

[22-Sep-2011 13:53:29] <cluther> Hackman238: Yeah. What's up?

[22-Sep-2011 13:53:32] <rmatte> humanix: Canada

[22-Sep-2011 13:54:08] <Hackman238> cluther: Someone was just telling me an epic Zenoss employee tale.

[22-Sep-2011 13:54:29] <Hackman238> cluther: About some theif who fell through a roof with a bag of money. ring a bell?

[22-Sep-2011 13:55:18] <Hackman238> cluther: Some people here over heard me bring up your name taking about zcml and the story telling began.

[22-Sep-2011 13:55:48] <cluther> Hackman238: I wish I could follow that reference, but.. I cannot.

[22-Sep-2011 13:56:12] <Hackman238> cluther: LOL alrighty.

[22-Sep-2011 13:57:09] <rmatte> Hackman238: An MSP that I worked for in the past (Nuvo Network Management) had someone try to break through a wall in to their server room.

[22-Sep-2011 13:57:39] <cluther> We did have a transient take some useless stuff out of our Annapolis office back before we had newfangled security devices like locked doors.

[22-Sep-2011 13:57:54] <cluther> He was well-dressed though because he had just come from stealing some suits.

[22-Sep-2011 13:57:55] <humanix> i see

[22-Sep-2011 13:58:02] <rmatte> haha

[22-Sep-2011 13:58:21] <Hackman238> rmatte: LOL

[22-Sep-2011 13:58:33] <Hackman238> cluther: LOL!!!

[22-Sep-2011 13:59:44] <rmatte> ...the server room was right beside the NOC area and had big windows, so the NOC guys who were working heard it right away, checked the security camera in the hallway, and called the police

[22-Sep-2011 13:59:46] <cluther> At my previous job the tenants on the floor above us had a fire that caused sprinklers to turn on in our "datacenter". Out of ~80 racks the only damaged equipment was a server named titanic.

[22-Sep-2011 14:00:18] <rmatte> It was one of those temporary type walls, so it was thin, they managed to make a pretty big hole before they realized that there were people in there and took off

[22-Sep-2011 14:00:40] <rmatte> cluther: haha, that's awesome

[22-Sep-2011 14:00:49] <Hackman238> cluther: Wow

[22-Sep-2011 14:01:41] <rmatte> sprinklers above racks, yowza

[22-Sep-2011 14:02:02] <Hackman238> rmatte: Must ahev well watered disks XD

[22-Sep-2011 14:02:35] <rmatte> but honestly, I'd be pretty pissed off if I were working in a datacenter and wasn't able to make it out of there in time before the gas went off

[22-Sep-2011 14:02:53] <rmatte> hopefully they have kill switches or something

[22-Sep-2011 14:03:03] <cluther> Hence the quotes around datacenter. It was crazy because we had a real fire suppression system in there.

[22-Sep-2011 14:03:17] <rmatte> lol

[22-Sep-2011 14:13:00] <Hackman238> cluther: LOL

[22-Sep-2011 14:13:22] <Hackman238> *was considering halon anti theft system for my car*

[22-Sep-2011 14:13:31] <Hackman238> "Its for fire supression"

[22-Sep-2011 14:13:37] <Hackman238> Bypass boobytrap laws

[22-Sep-2011 14:15:34] * Simon4 has been in a DC when the gas went off

[22-Sep-2011 14:15:37] <Simon4> I don't recommend it

[22-Sep-2011 14:15:40] <Simon4> no siree

[22-Sep-2011 14:17:26] <Hackman238> Simon4: Bad idea?

[22-Sep-2011 14:18:01] <Simon4> Hackman238: it was the non-lethal gas so you could still breathe etc - but it's fairly frightening and confusing/disorentating

[22-Sep-2011 14:18:19] <Simon4> it goes off with a hell of a bang

[22-Sep-2011 14:19:03] <Simon4> I was offered to stay in there with the engineers when they were live testing a new (empty) DC.. thought it was a great lark until it actually went off

[22-Sep-2011 14:19:30] <Hackman238> Simon4: Oh man

[22-Sep-2011 14:19:54] <Hackman238> Simon4: Yeah I was talking Halon 1301- not toxic, but in a car it'll displace all the O2 to 'supress' the theif

[22-Sep-2011 14:20:11] <Hackman238> Simon4: Either that or N2 gas

[22-Sep-2011 14:20:35] <Simon4> http://en.wikipedia.org/wiki/Inergen <-- was the stuff they use these days

[22-Sep-2011 14:20:47] <Simon4> you can still breathe it, but nothing will burn

[22-Sep-2011 14:20:49] <Hackman238> Simon4: N2 sublimizes though- other problems

[22-Sep-2011 14:21:26] <Hackman238> Simon4: Yeah but that wont kill the theif

[22-Sep-2011 14:21:31] <Hackman238> Simon4: Thats the whole idea

[22-Sep-2011 14:21:33] <Simon4> Hackman238: well no

[22-Sep-2011 14:21:34] <Hackman238> LOL

[22-Sep-2011 14:22:45] <Hackman238> Simon4: Hard to flood with just CO2  also- like N2 sublimizes and could plug exit nozzle, freeze stuff, etc

[22-Sep-2011 14:23:27] <Hackman238> Simon4: Though with CO2 the occupant will know they're dying. With N2 its a suprise when they realize they're about to pass out

[22-Sep-2011 14:23:55] <Hackman238> Simon4: Anyway...long and short of it is my legal staff told me dont do it. LOL

[22-Sep-2011 14:24:02] <Simon4> tehehe

[22-Sep-2011 14:27:12] <SDuensin> Your "legal staff"?

[22-Sep-2011 14:29:28] <Hackman238> SDuensin: Yeah, I have a team of lawyers on retainer

[22-Sep-2011 14:29:35] <Hackman238> Twelve of them

[22-Sep-2011 14:29:48] <SDuensin> Supreme Court?  :-)

[22-Sep-2011 14:29:51] <Hackman238> I like to push back

[22-Sep-2011 14:30:04] <Hackman238> No no, real lawyers

[22-Sep-2011 14:30:05] <Hackman238> LOL

[22-Sep-2011 14:30:09] <SDuensin> Got any WMI experts on retainer?

[22-Sep-2011 14:30:14] <Hackman238> Whitehall street, NYC

[22-Sep-2011 14:30:24] <Hackman238> SDuensin: BigEgor is the WMI man

[22-Sep-2011 14:30:35] * SDuensin thinks bigegor is bighiding

[22-Sep-2011 14:30:46] <Hackman238> SDuensin: Ryan might be able to help too

[22-Sep-2011 14:31:11] <SDuensin> I've got some awesome stuff going over here, but being able to monitor things might make it a bit more useful.  :-)

[22-Sep-2011 14:31:57] <Hackman238> SDuensin:

[22-Sep-2011 14:35:49] * SDuensin has suffered a segfault in module Templates and has been terminated.

[22-Sep-2011 14:36:45] <Hackman238> SDuensin: You've been terminated?

[22-Sep-2011 14:36:47] <Hackman238> LOL

[22-Sep-2011 14:37:24] <SDuensin> That entire Advanced->Monitoring Templates page makes zero sense to me.

[22-Sep-2011 14:37:37] <Hackman238> SDuensin: Yeah its a mess

[22-Sep-2011 14:38:32] <rmatte> SDuensin: click on the "Device Class" button at the bottom, it'll give you a more traditional view (should be the default view really)

[22-Sep-2011 14:39:42] <SDuensin> Whoa.  That almost makes sense.

[22-Sep-2011 14:40:28] <rmatte> yup

[22-Sep-2011 14:40:53] <rmatte> so what's not making sense at this point?

[22-Sep-2011 14:42:00] <SDuensin> Looking around still.

[22-Sep-2011 14:42:51] <SDuensin> I've made a tree of device types (did this a long time ago).  How do I get one assigned to the actual device?  All I see is /Devices.

[22-Sep-2011 14:44:21] <SDuensin> If I go to Bind Templates, I only see /Devices.

[22-Sep-2011 14:49:42] <Jane_Curry> Hi cluther:

[22-Sep-2011 14:51:54] <Jane_Curry> I just tried your overrides.zcml idea and I now get a large white Site Error page on login.  Can i just check...

[22-Sep-2011 14:55:08] <cluther> At least something happened.

[22-Sep-2011 14:57:01] <cluther> I tested it on 3.2.0 and it worked for me. Anything interesting in event.log? Actually, stop Zope and start it up in the foreground with "zopectl fg". Often this shows you your ZCML problem in an obvious way.

[22-Sep-2011 15:01:28] <Jane_Curry> cluther: Yeh - zope barfs.  I simply copied the lines from navigation.zcml as per the forum append and zope barfs I think because of the paths in there

[22-Sep-2011 15:02:32] <Jane_Curry> things like manager="../navigation.interfaces.IPrimaryNavigationMenu - I guess this is a relative path that barfs from my ZenPack location

[22-Sep-2011 15:02:50] <Jane_Curry> so I just need to figure either the correct relative path or an absolute path??

[22-Sep-2011 15:03:13] <cluther> This is what I used to test: https://gist.github.com/1235685

[22-Sep-2011 15:03:32] <cluther> The only difference was that I changed the permission to zenoss.ManageDMD so I could verify that "Reports" was gone for ZenUsers.

[22-Sep-2011 15:12:05] <Hackman238> jiveturkey: Welcome back

[22-Sep-2011 15:13:56] <straterra> rmatte: You're an op now? SAY WHAT'

[22-Sep-2011 15:21:10] <rmatte> I have been for some time

[22-Sep-2011 15:21:19] <straterra> You lie

[22-Sep-2011 15:21:27] <rmatte> lol

[22-Sep-2011 15:21:38] <Hackman238> LOL

[22-Sep-2011 15:21:52] <JohnnyNOC> hey guys.. i want to update a zenpack which i've cloned from git

[22-Sep-2011 15:22:14] <JohnnyNOC> par tof the install instructions say zenpack --link --install DIR

[22-Sep-2011 15:22:39] <JohnnyNOC> just wwant to confirm i can remove/move the DIR after installing it

[22-Sep-2011 15:22:58] <rmatte> JohnnyNOC: here's what I personally do, zip it up, rename the zip to .egg, install it like you normally would...

[22-Sep-2011 15:23:13] <JohnnyNOC> ah ok

[22-Sep-2011 15:23:22] <rmatte> then: cd in to the ZenPack directory in $ZENHOME/ZenPacks/ZenPacks.whatever.pack

[22-Sep-2011 15:23:42] <rmatte> then do: cp $ZENHOME/Products/ZenModel/ZenPackTemplate/* .

[22-Sep-2011 15:23:50] <rmatte> ignore the error about it not copying a directory

[22-Sep-2011 15:23:58] <rmatte> the ZenPack will then be in edit mode

[22-Sep-2011 15:24:10] <rmatte> you can then go in to the UI and change the version number, and whatever else

[22-Sep-2011 15:24:31] <rmatte> add whatever objects/files you want to it, then just export it like you normally would

[22-Sep-2011 15:25:18] <rmatte> It's really the simplest way to do it, no sense in messing around with linking and whatnot

[22-Sep-2011 15:25:35] <Hackman238> rmatte: That still workm in v3?

[22-Sep-2011 15:25:43] <Hackman238> rmatte: Doesnt work in v4

[22-Sep-2011 15:25:46] <rmatte> Hackman238: as far as I know, yeh

[22-Sep-2011 15:25:52] <Hackman238> rmatte: Good to know

[22-Sep-2011 15:25:56] <rmatte> why wouldn't it work in v4?

[22-Sep-2011 15:26:22] <cluther> JohnnyNOC: It is not safe to remove DIR after you've done a --link install from it.

[22-Sep-2011 15:26:22] <JohnnyNOC> that just seems overly complicated for updating/upgrading a zenpack :/

[22-Sep-2011 15:26:32] <JohnnyNOC> cluther i didn't think so, thank you

[22-Sep-2011 15:26:49] <rmatte> yeh, never remove a ZenPack directory before first uninstalling the ZenPack

[22-Sep-2011 15:26:50] <rmatte> lol

[22-Sep-2011 15:26:52] <JohnnyNOC> but would it be enough to zip it up and install it like i normally would?

[22-Sep-2011 15:26:54] <cluther> Honestly I like to install all ZenPacks using --link so I can live edit them.

[22-Sep-2011 15:26:56] <rmatte> you'll end up in a world of trouble

[22-Sep-2011 15:27:12] <Jane_Curry> cluther: Gotcha - changing the paths makes things much better

[22-Sep-2011 15:27:23] <rmatte> cluther: what's stopping you from live editing them in $ZENHOME/ZenPacks?

[22-Sep-2011 15:27:50] <cluther> rmatte: I can't commit my changes directly back to version control if they're installed from egg under $ZENHOME/ZenPacks/

[22-Sep-2011 15:27:54] <Jane_Curry> cluther: weird thing is that the overrides file is obviously taking effect cos when I hide it I go back to normal behaviour

[22-Sep-2011 15:28:08] <rmatte> cluther: ah

[22-Sep-2011 15:28:36] <Jane_Curry> cluther: but I left the permission the same at zenoss.Common so, in fact, nothing should change - I should see the REPORTS menu

[22-Sep-2011 15:28:45] <Jane_Curry> ... and I don't!!!

[22-Sep-2011 15:29:59] <JohnnyNOC> so the suggested way to install a zenpack after doing a git-clone is basically what rmatte suggested?

[22-Sep-2011 15:31:03] <JohnnyNOC> i guess i don't understand why the extra steps are needed.. :/

[22-Sep-2011 15:31:03] <cluther> JohnnyNOC: There are many ways, but I would suggest something simpler than what rmatte proposed.

[22-Sep-2011 15:31:22] <cluther> git clone ...

[22-Sep-2011 15:31:28] <cluther> zenpack --link --install <path>

[22-Sep-2011 15:31:32] <cluther> Get coffee.

[22-Sep-2011 15:31:53] <Hackman238> cluther: LOL

[22-Sep-2011 15:31:55] <JohnnyNOC> reason why i was wondering about that is because i've just git cloned to my homedir

[22-Sep-2011 15:32:05] <JohnnyNOC> and i don't want to have to leave that directory there, which is why i asked about removing it

[22-Sep-2011 15:32:06] <JohnnyNOC>

[22-Sep-2011 15:32:10] <Hackman238> cluther: Was just getting up to get coffee after setting off a zenoss script

[22-Sep-2011 15:32:12] <cluther> Do the git clone as the zenoss user.

[22-Sep-2011 15:32:31] <JohnnyNOC> i have

[22-Sep-2011 15:32:39] <cluther> You could even do so in the $ZENHOME/ZenPacks/ directory if you wanted to.

[22-Sep-2011 15:33:27] <JohnnyNOC> hm

[22-Sep-2011 15:33:35] <JohnnyNOC> but this is suppose dto be an update so that doesn't seem right

[22-Sep-2011 15:33:45] <JohnnyNOC> i trust you   it just doesn't seem right

[22-Sep-2011 15:35:10] <cluther> Update or not, it doesn't matter. Let's say you already have the ZenPack installed in .egg mode. Installing with --link from the git clone will work.

[22-Sep-2011 15:36:34] <JohnnyNOC> and can i remove the existing directory in $ZENHOME/ZenPacks?  or just ignore and leave it?

[22-Sep-2011 15:37:04] <dhopp> cluther:  I have a very basic zenpack that I did (basically just has some templates).  Because I'm a dummy I have blown away my dev environment that I built that zenpack with.  What you are saying is I could install the zenpack like I normally would..then do the zenpack —link magic and be able to edit it?

[22-Sep-2011 15:37:39] <cluther> The --link install (update) might actually clean that .egg directory up for you. If it doesn't, you can verify that the .egg directory no longer exists in $ZENHOME/ZenPacks/easy-install.pth after the update. If the .egg directory isn't in there anymore, you can safely delete the old .egg directory.

[22-Sep-2011 15:38:14] <cluther> dhopp: No, what you need to do is convert your ZenPack to development mode. This is where rmatte's instructions are applicable.

[22-Sep-2011 15:38:43] <JohnnyNOC> thanks cluther

[22-Sep-2011 15:39:21] <JohnnyNOC> i'll be sure to do a complete backup before tinkering around with this

[22-Sep-2011 15:39:56] <dhopp> cluther:  gotcha

[22-Sep-2011 15:40:52] <Jane_Curry> cluther: in fact, my reports menu has now disappeared for EVERYBODY even if I override the security permission to be zenoss.View

[22-Sep-2011 15:41:12] <rmatte> eugh, I asked enkrypt to update that OS Process issue ticket with more info and he didn't do it

[22-Sep-2011 15:41:23] <rmatte> oh well, his problem

[22-Sep-2011 15:45:13] <cluther> Jane_Curry: ok, I see that too.

[22-Sep-2011 15:47:53] <SDuensin> Does Zenoss reap "dead" Python processes running as "zenoss"?

[22-Sep-2011 15:48:50] <cluther> SDuensin: You'll probably have to be more specific if you're seeing something unusual. There's nothing in Zenoss that looks for processes owned by zenoss and kills them.

[22-Sep-2011 15:49:15] <cluther> Jane_Curry: I just noticed that in Zenoss 3.2 the Reports primary nav has the zenoss.Common permission by default.

[22-Sep-2011 15:49:49] <SDuensin> I wrote a script called "vnc-bouncer.py" that runs as a daemon.  After it's been running for awhile, it dies and the output says "Killed".  Nothing else on this box (SL6) except Zenoss.

[22-Sep-2011 15:50:13] <cluther> Check dmesg. Maybe OOMKiller.

[22-Sep-2011 15:50:42] <SDuensin> Good call!

[22-Sep-2011 15:51:09] <JohnnyNOC> cd

[22-Sep-2011 15:51:48] <SDuensin> Guess I have a leak in it somewhere.

[22-Sep-2011 16:16:56] <cluther> Jane_Curry: I haven't forgotten about you.. I'm plugging away at it.

[22-Sep-2011 16:18:42] <Jane_Curry> thanks cluther - I wanted to change the default from Common to View to hide it from those who don't have View (my specific users)

[22-Sep-2011 16:19:10] <Jane_Curry> but obviously I DO want others to see the REPORTS and most folk should have View

[22-Sep-2011 16:19:24] <cluther> Got it.

[22-Sep-2011 16:35:08] <humanix> i have a doubt

[22-Sep-2011 16:35:11] <humanix> about variable types

[22-Sep-2011 16:35:23] <humanix> im saving the connection numbers of a service to a gauge

[22-Sep-2011 16:35:30] <humanix> so i must see absolute values

[22-Sep-2011 16:35:31] <humanix> like

[22-Sep-2011 16:35:37] <humanix> 2 connections, 3 connections, 1 connection

[22-Sep-2011 16:35:41] <humanix> but i see

[22-Sep-2011 16:35:53] <humanix> 2.2 connectios on some part of my graph

[22-Sep-2011 16:36:00] <humanix> why does this happend?

[22-Sep-2011 16:36:36] <dhopp> humanix:  what time period are you looking at?

[22-Sep-2011 16:36:41] <dhopp> month? hour?

[22-Sep-2011 16:36:48] <humanix> maxium zoom

[22-Sep-2011 16:36:50] <humanix> 5 min interval

[22-Sep-2011 16:36:52] <dhopp> oh hrm

[22-Sep-2011 16:37:25] <cluther> humanix: Set the format of your graph points to %7.0lf

[22-Sep-2011 16:37:46] <cluther> RRDtool does interpolation between collected data points.

[22-Sep-2011 16:39:08] <rmatte> humanix: yeh, the default is 5.2lf%s, the .2 creates the decimal values, and the %s auto-converts to milli, kilo, giga, etc... If you make it 5.0lf or 7.0lf like cluther suggested you'll end up with absolute values

[22-Sep-2011 16:40:46] <humanix> graph deifnitions??

[22-Sep-2011 16:41:16] <humanix> or where do I set the format

[22-Sep-2011 16:41:33] <humanix> i see

[22-Sep-2011 16:41:34] <humanix> graph poin

[22-Sep-2011 16:41:42] <humanix> come like this %5.2lf%s

[22-Sep-2011 16:41:48] <humanix> i put %7.2lf%s ?

[22-Sep-2011 16:41:50] <humanix> sorry

[22-Sep-2011 16:41:54] <humanix> i put %7.0lf%s ?

[22-Sep-2011 16:42:27] <Hackman238> humanix: take off the %s

[22-Sep-2011 16:42:58] <Hackman238> humanix: Less you want kilo and mega connections

[22-Sep-2011 16:43:57] <rmatte> Hackman238: when I try to run that script you gave me it complains about the locale setting

[22-Sep-2011 16:44:06] <rmatte> Traceback (most recent call last):

[22-Sep-2011 16:44:06] <rmatte>   File "./util.py", line 14, in ?

[22-Sep-2011 16:44:06] <rmatte>     locale.setlocale(locale.LC_ALL, 'en_US')

[22-Sep-2011 16:44:06] <rmatte>   File "/usr/local/zenoss/python/lib/python2.4/locale.py", line 381, in setlocale

[22-Sep-2011 16:44:06] <rmatte>     return _setlocale(category, locale)

[22-Sep-2011 16:44:06] <rmatte> locale.Error: unsupported locale setting

[22-Sep-2011 16:45:13] <rmatte> will that script only work with Zenoss 3.x or something?

[22-Sep-2011 16:45:21] <Hackman238> rmatte: I use it in v2.5.2

[22-Sep-2011 16:45:30] <rmatte> that's the version I'm using it against

[22-Sep-2011 16:45:36] <Hackman238> rmatte: Looks like its the date locate

[22-Sep-2011 16:45:43] <Hackman238> *locale

[22-Sep-2011 16:46:21] <rmatte> I don't even get what that's used for

[22-Sep-2011 16:46:23] <Hackman238> rmatte: Just replace the locale.format commands

[22-Sep-2011 16:46:45] <Hackman238> rmatte: It gets the regional settings on the box and returns values that make sense for your country.

[22-Sep-2011 16:47:02] <Hackman238> rmatte: So in USA, 1,000.00 in Germany, 1.000,00

[22-Sep-2011 16:47:11] <rmatte> ah

[22-Sep-2011 16:47:24] <Hackman238> rmatte: No biggie. Just strip that crap out

[22-Sep-2011 16:47:26] <rmatte> I assume I can just replace it with print?

[22-Sep-2011 16:47:32] <Hackman238> Yep.

[22-Sep-2011 16:49:38] <rmatte> hmmm wtf

[22-Sep-2011 16:49:47] <rmatte> oh nevermind

[22-Sep-2011 16:50:27] <rmatte> ran it as my regular user account by accident

[22-Sep-2011 16:50:33] <Hackman238> rmatte: Got it? : )

[22-Sep-2011 16:50:46] <rmatte> yeh, seems ok now, now to pass it some options and test it

[22-Sep-2011 16:51:53] <Hackman238> rmatte: Gotcha

[22-Sep-2011 16:52:58] <rmatte> hmmm, doesn't seem to be working

[22-Sep-2011 16:53:12] <rmatte> ./util.py -s '09/01/2011 00:00:00' -e '09/21/2011 00:00:00'

[22-Sep-2011 16:53:15] <rmatte> returns nothing

[22-Sep-2011 16:53:54] <humanix> thanks guys!!!

[22-Sep-2011 16:54:56] <Hackman238> rmatte: Example /opt/zenoss/RQS.py -g '/CloudServers' -l '' -s '06/01/2011 00:00:00' -e '07/01/2011 00:00:00' -i 'Port-channel1'

[22-Sep-2011 16:55:11] <Hackman238> rmatte: use -l ''

[22-Sep-2011 16:55:29] <rmatte> still nothing

[22-Sep-2011 16:55:40] <Hackman238> rmatte: Also I forget if you need to specifiy what to run it against

[22-Sep-2011 16:55:53] <rmatte> hmmm, I want it to just run globally

[22-Sep-2011 16:56:03] <Hackman238> rmatte: try a group jsut for the hell of it

[22-Sep-2011 16:56:05] <rmatte> what I'm trying to do is this...

[22-Sep-2011 16:56:19] <Hackman238> rmatte: Its easy enough, jsut go in and chop out the condition checking in the for loop

[22-Sep-2011 16:56:20] <rmatte> Device Name | Average Device Util | Peak Device Util | Threshold Alerts

[22-Sep-2011 16:56:38] <rmatte> where the device utilization is the average percentage of utilization per interface

[22-Sep-2011 16:56:50] <rmatte> so I'll probably have to mod the script quite a bit

[22-Sep-2011 16:57:28] <Hackman238>     for d in dmd.Devices.getOrganizer(options.klass).getSubDevices():

[22-Sep-2011 16:57:28] <Hackman238>         if not d.monitorDevice(): continue

[22-Sep-2011 16:57:29] <Hackman238>         if not systemsMatch.search(d.getSystemNamesString()): continue

[22-Sep-2011 16:57:29] <Hackman238>         if not locationsMatch.search(d.getLocationName()): continue

[22-Sep-2011 16:57:29] <Hackman238>         for i in d.os.interfaces():

[22-Sep-2011 16:57:31] <Hackman238> #            if not intMatch.match(i.getInterfaceName()): continue

[22-Sep-2011 16:57:33] <Hackman238>             if i.getInterfaceName() != options.intfilter: continue

[22-Sep-2011 16:57:36] <Hackman238>             if i.getOperStatus()!= 1: continue

[22-Sep-2011 16:57:38] <Hackman238>             if not i.monitored(): continue

[22-Sep-2011 16:57:41] <Hackman238>             if i.snmpIgnore(): continue

[22-Sep-2011 16:58:03] <Hackman238> rmatte: Just commment out conditions that arent applicable and fix menu logic later

[22-Sep-2011 16:58:12] <rmatte> yup

[22-Sep-2011 16:58:32] <Hackman238> rmatte: Hagve to run. Good luck Msg me if you need help

[22-Sep-2011 16:58:35] <Hackman238> Later all!

[22-Sep-2011 17:00:11] <rmatte> later

[22-Sep-2011 17:02:15] <rmatte> I think my utilization script is closer to what I need than this one

[22-Sep-2011 17:02:23] <rmatte> since I'm looking for percentage value

[22-Sep-2011 17:02:26] <rmatte> which mine already does

[22-Sep-2011 17:02:57] <rmatte> this one is total traffic, which is cool, I should add that to mine, simple enough to do

[22-Sep-2011 17:27:27] tvincent_ is now known as tvincent

[22-Sep-2011 17:39:31] <JohnnyNOC> so i just updated this Zenpack i was asking about earlier and all seemed to go well except when i model a device it doesn't look like my modeler plugins are being run against the device

[22-Sep-2011 17:39:37] <JohnnyNOC> any ideas what may have gone wrong?

[22-Sep-2011 17:39:37] <JohnnyNOC> :/

[22-Sep-2011 17:40:35] <rmatte> JohnnyNOC: did you reinstall the pack from the linked directory after adding the plugin?

[22-Sep-2011 17:40:41] <rmatte> plugins get linked in to Zenoss upon install

[22-Sep-2011 17:41:00] <rmatte> you need to run an install against the linked directory again and then restart zope for your plugin changes to get picked up

[22-Sep-2011 17:41:12] <JohnnyNOC> ah, ok let me give that a shot

[22-Sep-2011 17:41:13] <JohnnyNOC> no i didn't do that

[22-Sep-2011 17:41:15] <rmatte> also make sure that the plugin is defined against the class or the device.

[22-Sep-2011 17:41:18] <JohnnyNOC> i did the zenpack --link --install

[22-Sep-2011 17:41:21] <JohnnyNOC> and all seemed to go well

[22-Sep-2011 17:41:25] <JohnnyNOC> and restarted zope

[22-Sep-2011 17:41:54] <rmatte> if it's a new plugin, have you linked it to the class or the device?

[22-Sep-2011 17:42:00] <JohnnyNOC> this is not new

[22-Sep-2011 17:42:04] <rmatte> k

[22-Sep-2011 17:42:09] <JohnnyNOC> it's supposed to be an update

[22-Sep-2011 17:42:14] <rmatte> then yeh, you need to reinstall after you make the changes

[22-Sep-2011 17:42:17] <rmatte> and restart zope

[22-Sep-2011 17:42:29] <rmatte> you might even have to restart Zenoss as a whole, but that's generally not necessary

[22-Sep-2011 17:42:41] <JohnnyNOC> no love

[22-Sep-2011 17:42:43] <JohnnyNOC> i got an error

[22-Sep-2011 17:42:48] <rmatte> what error?

[22-Sep-2011 17:43:21] <JohnnyNOC> hang on please

[22-Sep-2011 17:43:26] <rmatte> mhm

[22-Sep-2011 17:43:34] <JohnnyNOC> grr i have to stop doing this in screen, i can't scroll up

[22-Sep-2011 17:43:39] <rmatte> lol

[22-Sep-2011 17:43:48] <rmatte> there is a way to scroll in screen...

[22-Sep-2011 17:43:50] <JohnnyNOC> but i tried zenpack --install THENEWDIR and it complained

[22-Sep-2011 17:44:01] <JohnnyNOC> then i trid doing it to THENEWDIR.egg-link and it didn't work

[22-Sep-2011 17:44:14] <JohnnyNOC> but now i see THENEWDIR-VERSION

[22-Sep-2011 17:44:25] <JohnnyNOC> and i can't confirm that i saw that prior to trying to install them again cuz i can't scroll up

[22-Sep-2011 17:44:30] <JohnnyNOC> :[

[22-Sep-2011 17:44:46] <JohnnyNOC> this time it didn't complain however, so i'm restarting zenoss

[22-Sep-2011 17:46:34] <rmatte> aha, that's how you scroll up in screen, just remembered...

[22-Sep-2011 17:46:42] <rmatte> ctrl-a then escape

[22-Sep-2011 17:46:46] <rmatte> it'll put you in to copy mode

[22-Sep-2011 17:46:50] <rmatte> you can use page up and page down

[22-Sep-2011 17:46:55] <rmatte> then esc to return to normal

[22-Sep-2011 17:47:08] <JohnnyNOC> hah

[22-Sep-2011 17:47:08] <JohnnyNOC> nice

[22-Sep-2011 17:47:49] <JohnnyNOC> whatever i did, i'm kind of happy i see the version now

[22-Sep-2011 17:48:07] <rmatte> lol

[22-Sep-2011 17:48:07] <JohnnyNOC> and it looks like it picked up my plugins

[22-Sep-2011 17:48:17] <JohnnyNOC> does that make sense what i was saying earlier?

[22-Sep-2011 17:48:23] <JohnnyNOC> i git cloned a zenpack and the dir didn't include the version in it

[22-Sep-2011 17:48:51] <rmatte> yeh, makes sense

[22-Sep-2011 17:49:02] <rmatte> I've seen that before, I honestly don't get how they store it in git, it's weird

[22-Sep-2011 17:49:13] <rmatte> I always just take the regular .eggs and use those

[22-Sep-2011 17:51:09] <JohnnyNOC> it's not that i didn't want to take your advice earlier, but it seemed like way too many steps to try and upgrade a zenpack

[22-Sep-2011 17:51:10] <JohnnyNOC> :/

[22-Sep-2011 18:43:24] <rmatte> it's not even that many steps

[22-Sep-2011 18:43:27] <rmatte> it's like 3 steps

[22-Sep-2011 18:43:37] <rmatte> you actually probably took more steps doing it your way

[22-Sep-2011 18:43:37] <rmatte> lol

[22-Sep-2011 18:44:07] <rmatte> but then again, to have it in the git format that they expect I guess that way works better

[22-Sep-2011 18:53:21] <rmatte> well, time for me to sign off for the day, back tomorrow

[22-Sep-2011 18:53:26] <JohnnyNOC> thanks rmatte

[22-Sep-2011 18:53:30] <JohnnyNOC> have a great evening!

[22-Sep-2011 20:01:00] kocolosk is now known as kocolosk|away

[22-Sep-2011 21:37:33] <brockp> If I installed zenoss from a deb (from the dev.zenoss repo) is it safe to use zenpatch?

[22-Sep-2011 21:44:49] <brockp> When I updated from 3.1.0 to 3.2.0 zenoss claims now that the mysql server on my ubuntu machine is not running, even though It clearly is, It even shows up when I do an SNMP walk,

[22-Sep-2011 21:45:34] <brockp> When I compare it to a machine that does show its mysql running (gentoo) the difference is that the mysqld's that do not appear dont have an entry for HOST-RESOURCES-MIB::hrSWRunParameters.4292

[23-Sep-2011 00:00:01] [disconnected at Fri Sep 23 00:00:01 2011]

[23-Sep-2011 00:00:02] [connected at Fri Sep 23 00:00:02 2011]

[23-Sep-2011 00:00:11] [zenoss-logger (logger bot) has joined #zenoss]

[23-Sep-2011 08:20:04] <Hackman238> Morning all!

[23-Sep-2011 08:45:29] <JohnnyNOC> good morning

[23-Sep-2011 09:29:36] <Hackman238> http://www.livescience.com/16183-faster-speed-light-physics-breakthrough.html

[23-Sep-2011 09:36:35] <Jane_Curry> afternoon....

[23-Sep-2011 09:41:57] <JohnnyNOC> hey bigegor you around?

[23-Sep-2011 09:42:06] <JohnnyNOC> am i crazy with respec tto my latest comments on your zenpacks?

[23-Sep-2011 09:42:07] <JohnnyNOC>

[23-Sep-2011 09:47:44] <JohnnyNOC> sometimes i think i am

[23-Sep-2011 09:52:48] <Hackman238> This is crazy! - >http://www.engadget.com/2011/09/23/scientists-reconstruct-images-from-our-brains-plan-to-do-the-sa/

[23-Sep-2011 10:18:24] <humanix> hi there guys

[23-Sep-2011 10:21:01] <Hackman238> humanix: Hey

[23-Sep-2011 10:23:18] <humanix> how are you Hackman238

[23-Sep-2011 10:23:25] <humanix> @ office?

[23-Sep-2011 10:23:26] <humanix>

[23-Sep-2011 10:23:47] <Hackman238> humanix: I'm well.

[23-Sep-2011 10:23:50] <Hackman238> humanix: Yourself?

[23-Sep-2011 10:24:03] <humanix> me too

[23-Sep-2011 10:24:05] <humanix> @ office

[23-Sep-2011 10:25:10] <bigegor> hi all

[23-Sep-2011 10:28:17] <Hackman238> bigegor: Hey, how goes it?

[23-Sep-2011 10:28:38] <Hackman238> humanix: Why frown for being at the office?

[23-Sep-2011 10:30:29] <bigegor> Hackman238: it goes good

[23-Sep-2011 10:32:05] <brockp> Did things change for process monitoring in 3.2 from 3.1?  3.2 isn't seeing mysqld correctly

[23-Sep-2011 10:40:35] <SDuensin> FrIdAy!

[23-Sep-2011 10:46:45] <crazed> brockp: yeah

[23-Sep-2011 10:46:48] <crazed> they changed heavily

[23-Sep-2011 10:47:05] <crazed> i ran into a bunch of issues because of it

[23-Sep-2011 10:47:27] <brockp> Hmmm well a regex  for mysqld  wont find mysqld  on a server,

[23-Sep-2011 10:47:42] <brockp> It does find mysqld.bin on the same server though,

[23-Sep-2011 10:49:13] <humanix> ahahaha

[23-Sep-2011 10:49:17] <humanix> im used to Hackman238

[23-Sep-2011 10:49:24] <humanix> in fact, i like

[23-Sep-2011 10:49:33] <humanix> i assume a great part of us work @ a noc

[23-Sep-2011 10:52:24] <crazed> brockp: i can tell you that i've had no luck using .* type regular expressions

[23-Sep-2011 10:52:51] <brockp> my regex is exactly mysqld  no funny chars,

[23-Sep-2011 10:52:51] <crazed> things tend to work if you are more specific about each process you want monitored

[23-Sep-2011 10:53:08] <crazed> what's the process you want monitored

[23-Sep-2011 10:53:12] <crazed> and the regular expression

[23-Sep-2011 10:53:26] <brockp> mysql     1103  0.1  3.4 873944 70560 ?        Ssl  Sep22   0:54 /usr/sbin/mysqld

[23-Sep-2011 10:53:30] <brockp> regex is mysqld

[23-Sep-2011 10:53:59] <crazed> see if \/usr\/sbin\/mysqld fixes the problem

[23-Sep-2011 10:54:22] <brockp> Ok I will try that, but here is a weird thing, the regex is finding, the zenoss mysql server:

[23-Sep-2011 10:54:25] <brockp> mysql    15147  0.0  1.3 130292 28168 pts/1    Sl   10:09   0:01 /usr/local/zenoss/mysql/bin/mysqld.bin

[23-Sep-2011 10:55:07] <Hackman238> humanix: Gotcha.

[23-Sep-2011 10:55:21] <crazed> hm that is weird

[23-Sep-2011 10:56:23] <brockp> Yep, also finds mysqld on two gentoo servers,

[23-Sep-2011 10:56:55] <brockp> Still no luck,

[23-Sep-2011 10:56:58] <humanix> where are u from Hackman238 ?

[23-Sep-2011 10:57:46] <brockp> So strange thing is, that mysqld pit does not have an entry in HOST-RESOURCES-MIB::hrSWRunParameters  while all the mysqld's that it does find do

[23-Sep-2011 11:00:18] <brockp> I guess my statement about zenoss not finding it isn't quite true, if I remodel after I do the change, it

[23-Sep-2011 11:00:19] <brockp> Process not running: /usr/sbin/mysqld Using regex '\/usr\/sbin\/mysqld' All Processes have stopped since the last model occurred. Last Modification time (2011/09/23 10:56:11)

[23-Sep-2011 11:02:21] <crazed> 2011-09-23T10:15:00 WARNING ZODB.DB DB.open() has 9 open connections with a pool_size of 7

[23-Sep-2011 11:02:24] <crazed> i'm seeing that a lot

[23-Sep-2011 11:02:33] <crazed> any way to up the pool_size ?

[23-Sep-2011 11:05:27] <Hackman238> brockp: Ignore the pool messages

[23-Sep-2011 11:05:49] <Hackman238> brockp: Even if you add pool-size 50 to zope.conf those messages wills till show. No idea why

[23-Sep-2011 11:06:24] <Hackman238> crazed: Rather I meant that to you crazed

[23-Sep-2011 11:07:18] <crazed> ok thanks

[23-Sep-2011 11:07:33] <crazed> Hackman238: any idea what would cause an enormous amount of unable to read processes alerts in 3.2?

[23-Sep-2011 11:07:43] <crazed> i never had these in 3.1

[23-Sep-2011 11:09:04] <Hackman238> crazed: Fail toread on linux, unix or windows boxes?

[23-Sep-2011 11:09:16] <crazed> Hackman238: linux

[23-Sep-2011 11:09:21] <crazed> we don't have any windows thankfully

[23-Sep-2011 11:10:27] <Hackman238> crazed: LOL

[23-Sep-2011 11:10:51] <Hackman238> crazed: There are a number of patched for v3.2, there might be one for that problem

[23-Sep-2011 11:12:35] <crazed> i've been going through

[23-Sep-2011 11:12:51] <brockp> How do we find what patches are available for  a release? Other than the ones in the subject of this channel?

[23-Sep-2011 11:13:03] <crazed> Hackman238: i've been going through http://dev.zenoss.org/trac/query?status=assigned&status=backlog&status=closed&status=fixrejected&status=new&status=verification&group=patch_state&patch=3.2.1&col=id&col=summary&col=status&col=owner&col=priority&col=resolution&col=deployed&col=changetime&report=6&order=status

[23-Sep-2011 11:13:10] <crazed> is that the right place for these patches?

[23-Sep-2011 11:13:44] kocolosk|away is now known as kocolosk

[23-Sep-2011 11:14:30] <humanix> Hackman238, where are u from?

[23-Sep-2011 11:26:14] <Hackman238> humanix: New York, USA

[23-Sep-2011 11:27:10] <Hackman238> crazed: Should be

[23-Sep-2011 11:33:34] <humanix> i see

[23-Sep-2011 11:33:56] <Hackman238> humanix: I'm in Texas, USA now.

[23-Sep-2011 11:34:01] <Hackman238> humanix: Horrible weather here.

[23-Sep-2011 11:34:05] <Hackman238> : )

[23-Sep-2011 11:34:20] <humanix> you live in texas now?

[23-Sep-2011 11:34:25] <humanix> or you are cause of work?

[23-Sep-2011 11:35:13] <Hackman238> humanix: I live in Texas now for work at Rackspace. Still ahve my business back in New York and on in California up and going.

[23-Sep-2011 11:35:29] <Hackman238> s/on//

[23-Sep-2011 11:35:39] <humanix> haha

[23-Sep-2011 11:35:42] <humanix> i see

[23-Sep-2011 11:35:50] <Hackman238> humanix: Sorry, typing and speaking in a confrence simultaniously

[23-Sep-2011 11:35:58] <humanix> its ok

[23-Sep-2011 11:36:04] <humanix> im from Santiago, Chile

[23-Sep-2011 11:36:24] <Hackman238> humanix: Very cool

[23-Sep-2011 11:37:20] <humanix> haha

[23-Sep-2011 11:37:22] <humanix> far away

[23-Sep-2011 11:37:44] <Hackman238> humanix: Indeed.

[23-Sep-2011 11:42:54] <crazed> damn none of the patches fixed the unable to read processes issue

[23-Sep-2011 11:47:55] <brockp> crazed: I think you might see the same problem I am seeing,

[23-Sep-2011 11:49:00] <crazed> i do have a bunch of process issues in general

[23-Sep-2011 11:49:15] <brockp> Does 3.2  find some processes but not others that are actually running?

[23-Sep-2011 11:49:25] <crazed> it uses md5 hashes

[23-Sep-2011 11:49:28] <crazed> i believe that's new

[23-Sep-2011 11:49:36] <crazed> they wanted to match the way the modler finds processes

[23-Sep-2011 11:50:18] <brockp> snmpwalk -v 1 -c public HOSTNAME hrSWRunTable  | grep processes

[23-Sep-2011 11:51:00] <brockp> If you run that are the ones that are found have an entry with hrSWRunParameters  and the ones that are not missing that?

[23-Sep-2011 11:51:17] <crazed> hm let me check

[23-Sep-2011 11:54:12] <crazed> brockp: nope it definitely shows up there

[23-Sep-2011 11:54:24] <crazed> i've had good success by removing all OS processes and remodeling

[23-Sep-2011 11:54:27] <brockp> hmm, ok maybe my hunch was bad then,

[23-Sep-2011 11:54:28] <crazed> that tends to fix the problem

[23-Sep-2011 11:54:49] <brockp> hmmm maybe I will try that,  you go into the host and just delete them all?

[23-Sep-2011 11:55:16] <crazed> yeah

[23-Sep-2011 11:55:30] <crazed> kinda blows when you have 100s of nodes

[23-Sep-2011 11:55:33] <crazed> doing the same thing

[23-Sep-2011 11:57:16] <brockp> yeah I only have one where it doesn't find a single mysql it used to find,

[23-Sep-2011 11:57:30] <brockp> I may need to roll back to 3.1 because this is buggy as hell in 3.2

[23-Sep-2011 12:00:01] [disconnected at Fri Sep 23 12:00:01 2011]

[23-Sep-2011 12:00:02] [connected at Fri Sep 23 12:00:02 2011]

[23-Sep-2011 12:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[23-Sep-2011 12:00:16] <crazed> poor choice, because 3.2 definitely isn't ready

[23-Sep-2011 12:00:52] <brockp> good thing I have bacula running zenbackup every night,

[23-Sep-2011 12:01:19] <crazed> yeah ... i was under the impression my backups were working

[23-Sep-2011 12:01:33] <crazed> but we migrated to a bigger zenoss server

[23-Sep-2011 12:01:38] <crazed> and forgot to migrate backups

[23-Sep-2011 12:04:10] <brockp> ouch, good luck,

[23-Sep-2011 12:08:44] <crazed> Unable to read processes on device HOST; error: www_bin_httpd b29e3ebc96638a67baac706c2fc5ecb4

[23-Sep-2011 12:08:48] <crazed> like i don't even know what that menas

[23-Sep-2011 12:20:31] <jbk> hmm.. is there any way to restrict who can run certain commands in zenoss?

[23-Sep-2011 12:25:20] <Hackman238> jbk: In core theres only all or nothing

[23-Sep-2011 12:25:36] <Hackman238> jbk: Jane has been working on a ACL pack to adress user permisisons

[23-Sep-2011 12:25:39] <JohnnyNOC> 

[23-Sep-2011 12:26:07] <Hackman238> crazed: Can you delete the processes, clear the events and remodel?

[23-Sep-2011 12:26:22] <SDuensin> Core vs. Enterprise is getting annoying.  :-/

[23-Sep-2011 12:26:32] <Hackman238> SDuensin: Ent hasn't ACL either

[23-Sep-2011 12:26:38] <Hackman238> SDuensin: **useful ACL

[23-Sep-2011 12:26:42] <SDuensin> lol

[23-Sep-2011 12:26:53] <Hackman238> SDuensin:

[23-Sep-2011 12:27:12] <Hackman238> SDuensin: Its being addressed. In v4.0.2 the ACL's are better, but they still suck.

[23-Sep-2011 12:27:37] <Hackman238> SDuensin: In general, however, only engineers/noc should see Zenoss not non tech people

[23-Sep-2011 12:27:43] <SDuensin> Hackman238, I'm going to go off the deep end here in a minute.  Investigating writing a WMI module that actually works.  I have the Windows side done - now I need the Zenoss side.

[23-Sep-2011 12:27:51] <Hackman238> SDuensin: Use reportmail to schedule reports to customers who want graphs

[23-Sep-2011 12:27:56] <crazed> Hackman238: yes, that tends to fix the problem

[23-Sep-2011 12:28:08] <crazed> unfortunately i have about 50 of them doing this

[23-Sep-2011 12:28:16] <crazed> is there a cleaner way to do this in bulk?

[23-Sep-2011 12:28:26] <Hackman238> crazed: Gotcha. In ZenDMD, but lets try something

[23-Sep-2011 12:28:33] <Hackman238> crazed: su zenoss; zenchkrels -x1 -r

[23-Sep-2011 12:28:40] <Hackman238> crazed: Sounds like those objs are funky

[23-Sep-2011 12:28:41] <crazed> what exactly does that do

[23-Sep-2011 12:28:52] <Hackman238> crazed: It checks relationships and rebuilds them when broken

[23-Sep-2011 12:28:57] <Hackman238> crazed: Very safe4

[23-Sep-2011 12:29:10] <Hackman238> crazed: Avoid zenfixit.py, however. Very very dangerous

[23-Sep-2011 12:29:29] <crazed> okay

[23-Sep-2011 12:29:29] <Hackman238> SDuensin: Whats the obsitcal? Maybe I can help

[23-Sep-2011 12:29:33] <crazed> sounds safe enough for me

[23-Sep-2011 12:29:50] <SDuensin> Hackman238, WMI doesn't work.  :-)

[23-Sep-2011 12:29:51] <Hackman238> crazed: Try to keep up with backups though

[23-Sep-2011 12:30:10] <Hackman238> SDuensin: Wheres it failing now?

[23-Sep-2011 12:30:29] <SDuensin> Hackman238, I had hoped the WMI data source used the 'wmic' tool to fetch data.  Doesn't seem the ZenPack for WMI uses it.  That'd have been easy.

[23-Sep-2011 12:31:17] <Hackman238> SDuensin: Ah gotcha. I think it uses a WMI module

[23-Sep-2011 12:31:34] <SDuensin> Yea.  If I could swap out the wmic binary, I'd be golden.

[23-Sep-2011 12:31:45] <Hackman238> SDuensin: You could use zencommand

[23-Sep-2011 12:31:46] <crazed> Hackman238: of course i plan on making sure backups work

[23-Sep-2011 12:32:00] <crazed> waiting for this zenchkrels to finish, i have a lot of nodes in there

[23-Sep-2011 12:32:04] <SDuensin> Hackman238, looking into that now.

[23-Sep-2011 12:32:10] <Hackman238> crazed: Try to avoid zenbackup too- just zeopack and copy out the data.fs.old

[23-Sep-2011 12:32:12] <crazed> ~600 or so

[23-Sep-2011 12:32:18] <crazed> oh really?

[23-Sep-2011 12:32:24] <crazed> i was going to use zenbackup

[23-Sep-2011 12:32:27] <Hackman238> crazed: Yeah that will be the unpacked zope

[23-Sep-2011 12:32:39] <Hackman238> crazed: If you want a packed backup, run zeopack twice

[23-Sep-2011 12:32:51] <crazed> hm

[23-Sep-2011 12:32:55] <crazed> i'll have to look into that

[23-Sep-2011 12:32:59] <Hackman238> crazed: The data.fs is all of zope. Then just make a tar copy of zenoss once in a while

[23-Sep-2011 12:33:23] <Hackman238> crazed: for the mysql, it depends on how important it is- just dump it once and a while for safe keeping.

[23-Sep-2011 12:33:40] <crazed> tar up /opt/zenoss once in awhile?

[23-Sep-2011 12:33:53] <Hackman238> crazed: RRD backups are tricky since so many files- up to you how to handle

[23-Sep-2011 12:33:54] <crazed> i mean we need to make sure to keep the rrd files since those are important to us

[23-Sep-2011 12:33:58] <crazed> yeah

[23-Sep-2011 12:34:01] <crazed> it is tricky

[23-Sep-2011 12:34:50] <Hackman238> crazed: what we do for rrd's is keep a backup on spindle array using threaded copy from the storage array thats an ssd raid 10

[23-Sep-2011 12:35:04] <Hackman238> crazed: No other way to backup millions for files quickly and safely that I've found

[23-Sep-2011 12:35:13] <Hackman238> *s/for/of/

[23-Sep-2011 12:36:58] <crazed> hm agreed

[23-Sep-2011 12:37:02] <LiamMac> Anyone know of a good way to export a subset of devices on 1 ZenOSS install and then import those devices into their own Standalone ZenOSS instance?

[23-Sep-2011 12:37:06] <crazed> i don't think i'll ever get that kind of hardware for monitoring though

[23-Sep-2011 12:37:19] <Hackman238> LiamMac: zendevicedump

[23-Sep-2011 12:37:23] <Hackman238> LiamMac: zendeviceload

[23-Sep-2011 12:37:35] <LiamMac> nice thanks Hackman238

[23-Sep-2011 12:37:36] <Hackman238> crazed: At a large enough scale you'll need it

[23-Sep-2011 12:37:42] <Hackman238> LiamMac: NP

[23-Sep-2011 12:39:09] <jbk> is there a way to restrict who can run certain commands for a device?

[23-Sep-2011 12:39:39] <Hackman238> jbk: restrice the commands menu per device?

[23-Sep-2011 12:39:45] <Hackman238> jbk: Not easily.

[23-Sep-2011 12:41:10] <Hackman238> jbk: You can add context sensitive menu entries, but I dont think the command  menu does that

[23-Sep-2011 12:41:39] <Hackman238> jbk: E.g. menu entry thats only visible when the object class equals something

[23-Sep-2011 12:44:41] <crazed> Hackman238: http://pastebin.com/UFpYzrjx

[23-Sep-2011 12:44:49] <crazed> that's what i got during the chkrels

[23-Sep-2011 12:44:57] <Hackman238> http://money.cnn.com/video/technology/2011/09/13/f_bctwf_rackspace.fortune/

[23-Sep-2011 12:48:08] <Hackman238> crazed: su zenoss; zendmd

[23-Sep-2011 12:48:14] <Hackman238> crazed: run: for user in dmd.ZenUsers.getUsers(): user.messages.removeRelation()

[23-Sep-2011 12:48:20] <Hackman238> crazed: commit()

[23-Sep-2011 12:48:26] <Hackman238> crazed: CTRL + D

[23-Sep-2011 12:48:31] <Hackman238> crazed: zenchkrels -x1 -r

[23-Sep-2011 12:50:10] <crazed> getUsers?

[23-Sep-2011 12:50:17] <crazed> not sure what users have to do in this case

[23-Sep-2011 12:50:25] <crazed> just wondering what exactly i'm changing here haha

[23-Sep-2011 12:50:30] <Hackman238> crazed: messageQueues are corrupted

[23-Sep-2011 12:50:41] <Hackman238> crazed: Prevents us from completeing zenchkrels

[23-Sep-2011 12:50:43] <Hackman238>

[23-Sep-2011 12:50:51] <crazed> hmm

[23-Sep-2011 12:50:56] <crazed> and it doesn't matter if we blow them away?

[23-Sep-2011 12:51:07] <Hackman238> crazed: They'll be rebuilt

[23-Sep-2011 12:51:14] <crazed> cool

[23-Sep-2011 12:52:45] <crazed> thanks, lets see where this gets me

[23-Sep-2011 12:52:55] <crazed> zenchkrels is running, probably take another few minutse

[23-Sep-2011 12:57:11] <Hackman238> crazed: Alrighty

[23-Sep-2011 12:57:27] <crazed> Hackman238: hmm same error

[23-Sep-2011 12:57:56] <Hackman238> crazed: When you ran that zendmd code, did you do a commit() when it finished?

[23-Sep-2011 12:58:09] <crazed> yeah

[23-Sep-2011 12:58:16] <Hackman238> crazed: No errors?

[23-Sep-2011 12:58:24] <crazed> nope i didn't see any

[23-Sep-2011 12:58:41] <Hackman238> crazed: yikes.

[23-Sep-2011 12:58:46] <Hackman238> crazed: lets do it a second time

[23-Sep-2011 12:58:50] <crazed> okay

[23-Sep-2011 12:58:52] <Hackman238> crazed: su zenoss; zendmd

[23-Sep-2011 12:59:10] <Hackman238> for user in dmd.ZenUsers.getUsers():

[23-Sep-2011 12:59:15] <Hackman238> {enter}

[23-Sep-2011 12:59:23] <crazed> tab

[23-Sep-2011 12:59:28] <crazed> yeah i know how to use python's shell

[23-Sep-2011 12:59:33] <Hackman238> {4 spaces}user.messages.removeRelation()

[23-Sep-2011 12:59:36] <Hackman238> {double enter}

[23-Sep-2011 12:59:41] <Hackman238> crazed: Sorry

[23-Sep-2011 13:00:37] <crazed> i ran it again, so let's see what happens

[23-Sep-2011 13:03:24] <Hackman238> crazed: could also try a=dmd.ZenUsers.getUser('admin') a.messageQueue.removeRelation()

[23-Sep-2011 13:03:31] <Hackman238> crazed: could also try a=dmd.ZenUsers.getUser('admin') a.messages.removeRelation()

[23-Sep-2011 13:04:23] <crazed> we use ldap users

[23-Sep-2011 13:04:25] <crazed> if that matters

[23-Sep-2011 13:04:33] <Hackman238> crazed: Shouldn't

[23-Sep-2011 13:05:06] <Hackman238> crazed: Need coffee. BRB

[23-Sep-2011 13:05:14] <crazed> AttributeError: 'NoneType' object has no attribute 'messageQueue'

[23-Sep-2011 13:05:23] <crazed> coffee! good idea

[23-Sep-2011 13:05:39] <Hackman238> crazed: if you do just a

[23-Sep-2011 13:05:49] <Hackman238> crazed: does it return propertied user admin?

[23-Sep-2011 13:08:56] <crazed> >>> a=dmd.ZenUsers.getUser('admin')

[23-Sep-2011 13:08:56] <crazed> >>> a

[23-Sep-2011 13:08:56] <crazed> >>>

[23-Sep-2011 13:09:12] <Hackman238> crazed: Oh wow

[23-Sep-2011 13:09:22] <Hackman238> crazed: your shit's borked

[23-Sep-2011 13:09:25] <crazed> lol

[23-Sep-2011 13:09:26] <Hackman238> crazed: Hum.

[23-Sep-2011 13:09:48] <Hackman238> >>> a=dmd.ZenUsers.getUser('admin')

[23-Sep-2011 13:09:48] <Hackman238> >>> a

[23-Sep-2011 13:09:48] <Hackman238> <PropertiedUser 'admin'>

[23-Sep-2011 13:09:48] <Hackman238> >>>

[23-Sep-2011 13:10:05] <crazed> i see a bunch of users

[23-Sep-2011 13:10:07] <crazed> if i do getusers

[23-Sep-2011 13:10:15] <Hackman238> crazed: is admin in the list?

[23-Sep-2011 13:10:29] <crazed> nope

[23-Sep-2011 13:10:46] <Hackman238> crazed: try a dmd.ZenUsers.reindex_all()

[23-Sep-2011 13:10:50] <Hackman238> crazed: then commit()

[23-Sep-2011 13:11:06] <Hackman238> crazed: from the web UI, add a new user named admin

[23-Sep-2011 13:11:20] <Hackman238> crazed: Then in dmd do sync()

[23-Sep-2011 13:11:25] <crazed> we probably deleted the admin user

[23-Sep-2011 13:11:27] <crazed> at some point

[23-Sep-2011 13:11:31] <Hackman238> crazed: and retry the getUser('admin')

[23-Sep-2011 13:11:42] <Hackman238> crazed: Cant- its needed for all sorts of processes

[23-Sep-2011 13:11:52] <Hackman238> crazed: Its just semi removed.

[23-Sep-2011 13:12:28] <crazed> hm

[23-Sep-2011 13:12:30] <crazed> admin exists in the ui

[23-Sep-2011 13:12:31] <crazed> actually

[23-Sep-2011 13:12:34] <Hackman238> crazed: Alrighty, brb. Coffee for real this time

[23-Sep-2011 13:18:18] <Hackman238> crazed: Hum

[23-Sep-2011 13:18:38] <crazed> yeah

[23-Sep-2011 13:18:51] <crazed> so i guess icould delete/readd but i don't know if that's the best action

[23-Sep-2011 13:19:03] <Hackman238> crazed: it should stop you deleting admin in the ui

[23-Sep-2011 13:19:22] <Hackman238> crazed: try for i in dmd.ZenUsers.getSubObjects():

[23-Sep-2011 13:19:29] <Hackman238> crazed:    i.id

[23-Sep-2011 13:19:33] <Hackman238> crazed: Admin in that list?

[23-Sep-2011 13:20:01] <crazed> Hackman238: yes

[23-Sep-2011 13:20:28] <Hackman238> try a=dmd.ZenUsers._getOb('admin')

[23-Sep-2011 13:20:42] <Hackman238> then a

[23-Sep-2011 13:20:49] <crazed> <UserSettings at /zport/dmd/ZenUsers/admin>

[23-Sep-2011 13:20:51] <crazed> is what i get

[23-Sep-2011 13:20:57] <Hackman238> crazed: nice

[23-Sep-2011 13:21:10] <Hackman238> a.messages.removeRelation()

[23-Sep-2011 13:21:13] <Hackman238> commit()

[23-Sep-2011 13:22:10] <crazed> nice

[23-Sep-2011 13:22:13] <crazed> and now the zenchkrel?

[23-Sep-2011 13:22:20] <Hackman238> crazed: yes please

[23-Sep-2011 13:28:06] <crazed> Hackman238: it finished successfully

[23-Sep-2011 13:28:24] <Hackman238> crazed: Nice

[23-Sep-2011 13:28:29] <Hackman238> restart zenprocess

[23-Sep-2011 13:28:43] <Hackman238> crazed: and clear the failure events

[23-Sep-2011 13:28:55] <Hackman238> crazed: Lets see if the problem was relationships being borked

[23-Sep-2011 13:29:04] <Hackman238> crazed: could be other things too

[23-Sep-2011 13:30:31] <crazed> Hackman238: oh no now i'm getting a bunch of process not running

[23-Sep-2011 13:30:52] <bigegor> if someone want monitor Storage with SMI-S agent, i've made ZenPack for that. It must work with ANY SMI-S compatible agents. here is link: https://github.com/epuzanov/ZenPacks.community.SMISMon

[23-Sep-2011 13:32:00] <crazed> actually it seems to only be a few currently

[23-Sep-2011 13:32:46] <Hackman238> crazed: I dont understand- did it improve or worsen?

[23-Sep-2011 13:32:55] <Hackman238> bigegor: Thanks egor!

[23-Sep-2011 13:33:02] <crazed> Hackman238: definitely improved

[23-Sep-2011 13:33:04] <bigegor> for now SMISMon ZenPack not ready for Productions environment (BETA)

[23-Sep-2011 13:33:44] <Hackman238> crazed: If it fails, we can delete them ALL for recreation

[23-Sep-2011 13:33:47] <Hackman238> crazed: zendmd

[23-Sep-2011 13:33:58] <Hackman238> crazed: for d in dmd.Devices.deviceSearch():

[23-Sep-2011 13:34:07] <Hackman238>     for p in d.os.processes():

[23-Sep-2011 13:34:18] <Hackman238>         p.manage_deleteComponent()

[23-Sep-2011 13:34:25] <Hackman238> {double enter}

[23-Sep-2011 13:34:27] <Hackman238> commit()

[23-Sep-2011 13:34:40] <Hackman238> It'll delete them all

[23-Sep-2011 13:34:48] <Hackman238> bigegor: Good to know

[23-Sep-2011 13:35:30] <bigegor> i've tested it against HP EVA, feedback about other storages types highly appreciated

[23-Sep-2011 13:35:54] <Hackman238> bigegor: I'll pass it to our SAN team to see if they want to test

[23-Sep-2011 13:35:57] <crazed> Hackman238: delete what all?

[23-Sep-2011 13:36:10] <Hackman238> crazed: all the processes on each device

[23-Sep-2011 13:36:16] <Hackman238> crazed: Last ditch effort to fix

[23-Sep-2011 13:36:23] <crazed> ah yeah i'm going to avvoid that for now

[23-Sep-2011 13:36:35] <Hackman238> crazed: Save the code just in case

[23-Sep-2011 13:36:56] <crazed> will do

[23-Sep-2011 13:37:37] <Hackman238> Alrighty

[23-Sep-2011 13:40:16] <crazed> there was only about 5 nodes

[23-Sep-2011 13:40:19] <crazed> i'm just remodeling them

[23-Sep-2011 13:40:26] <crazed> and it seems to be clearing my issues

[23-Sep-2011 13:40:36] <Hackman238> crazed: Very cool

[23-Sep-2011 13:59:06] <jbk> can you create an event mapping that parses one of the bits of data fron an SNMP trap to (for example) set the component?

[23-Sep-2011 13:59:20] <jbk> i can't find any examples offhand

[23-Sep-2011 14:12:16] <Hackman238> if evt.message == 'snmp trap authenticationFailure': evt.summary = "SNMP Authentication Failure from " + getattr(evt,"1.3.6.1.4.1.9.2.1.5.0") evt.eventClass = "/Security/Auth"

[23-Sep-2011 14:12:33] <Hackman238> jbk: the getattr(evt,oid) is what you're looking for I think

[23-Sep-2011 14:15:07] <jbk> hopefully i can release this publically, but i'm revamping the Netapp zenpack

[23-Sep-2011 14:15:12] <jbk> though got a few annoying issues to figure out

[23-Sep-2011 14:15:22] <jbk> (the netapp mib could be better)

[23-Sep-2011 14:16:05] <Hackman238> jbk: Very cool

[23-Sep-2011 14:16:21] <jbk> for zEventClearClasses, should it be the full path /Foo/Bar/Baz of the event class?

[23-Sep-2011 14:16:55] <jbk> there's also i think a bug in the FileSystem.py code -- some of the block sizes are still returning int's

[23-Sep-2011 14:17:01] <jbk> would like to submit a patch for that

[23-Sep-2011 14:17:23] <Hackman238> jbk: shouldnt block size be int?

[23-Sep-2011 14:17:31] <jbk> long would be better

[23-Sep-2011 14:17:34] <Hackman238> jbk: yes to the event class

[23-Sep-2011 14:17:41] <Hackman238> jbk: Ah true true

[23-Sep-2011 14:17:42] <jbk> not the size of the a block

[23-Sep-2011 14:17:47] <jbk> but the number of blocks

[23-Sep-2011 14:17:54] <Hackman238> jbk: Right

[23-Sep-2011 14:18:30] * jbk got a few negative values

[23-Sep-2011 14:19:22] <Hackman238> jbk: Always a win for disk size

[23-Sep-2011 14:19:27] <Hackman238> jbk: LOL

[23-Sep-2011 14:19:30] <Hackman238>

[23-Sep-2011 14:23:13] <jbk> how can i match the component value for an event?

[23-Sep-2011 14:23:29] <jbk> for example, if a volume goes online

[23-Sep-2011 14:23:46] <jbk> an oid 'productTrapData' contains the string 'Volume xxxxx is online.'

[23-Sep-2011 14:23:49] <Hackman238> jbk: evt.component

[23-Sep-2011 14:24:04] <jbk> i can use python to pull the volume name from the string (evt.productTrapData)

[23-Sep-2011 14:24:29] <jbk> well is that a free form string, or a reference to an actual class

[23-Sep-2011 14:24:49] <jbk> i.e. does it need to be '/Storage/NetApp/filername/filesystemname'

[23-Sep-2011 14:25:05] <Hackman238> jbk: just the mount name

[23-Sep-2011 14:25:17] <Hackman238> jbk: must equal what ${here/mount} would return

[23-Sep-2011 14:26:59] <jbk> ok, another question

[23-Sep-2011 14:27:21] <jbk> pretty much all netapp traps follow that pattern (return a string in productTrapData)

[23-Sep-2011 14:27:48] <jbk> can i leave the class key blank to have a mapping that sets the summary from that for all the traps?

[23-Sep-2011 14:27:52] <jbk> or do i have to do it for each one?

[23-Sep-2011 14:28:18] <jbk>  /Storage/NetApp <-- have transform to set evt.summary

[23-Sep-2011 14:28:42] <jbk>  /Storage/NetApp/Volume/foo <- have specific trap (w/ classkey) w/ possible additional transformations

[23-Sep-2011 14:31:42] <cluther> jbk: Mappings with "defaultmapping" as their event class key will be applied as a cleanup rule to events with eventClassKeys that don't match specific mappings or with blank eventClassKeys.

[23-Sep-2011 14:32:35] <cluther> jbk: So create a default mapping and use the rule field to verify the existence of the productTrapData. Then you can handle all of these traps in one transform.

[23-Sep-2011 14:32:57] <jbk> but that will only get applied if no class key exists, right?

[23-Sep-2011 14:33:13] <jbk> i'm trying to avoid having to add 'evt.summary = evt.productTrapData' for every defined event class

[23-Sep-2011 14:33:15] <cluther> It'll also get applied if no other other mappings match the specific eventClassKey.

[23-Sep-2011 14:33:19] <jbk> basically

[23-Sep-2011 14:34:20] <cluther> jbk: It might be even easier to use the hierarchical application of transforms.

[23-Sep-2011 14:34:33] <cluther> Create a transform on the root event class that does..

[23-Sep-2011 14:35:07] <cluther> if getattr(evt, 'productTrapData', None): evt.summary = evt.productTrapData

[23-Sep-2011 14:35:28] <cluther> It'll be applied for absolutely every event that comes into the system.

[23-Sep-2011 14:35:47] <cluther> Of course you could move this down to a more specific event class if that's appropriate.

[23-Sep-2011 14:36:56] <jbk> well if i create it at a higher level, but not the top

[23-Sep-2011 14:37:32] <jbk> once it finds a class key that matches, does it apply any other class mappings at higher levels of the heirarchy?

[23-Sep-2011 14:37:48] <jbk> for example, if i create a mapping at /Foo/Bar w/ a blank class key

[23-Sep-2011 14:37:57] <jbk> then one at /Foo/Bar/Baz with a class key

[23-Sep-2011 14:38:27] <jbk> will it get set to /Foo/Bar/Baz, then apply the blank one in /Foo/Bar ?

[23-Sep-2011 14:38:49] <cluther> jbk: No, only one mapping will ever get applied to an event.

[23-Sep-2011 14:39:14] <cluther> However, all of the event class transforms will run on the event.

[23-Sep-2011 14:40:29] <jbk> ok i think i asked it wrong, but i think it'll work

[23-Sep-2011 14:40:41] <jbk> I create a /Storage/NetApp/Volume class

[23-Sep-2011 14:41:04] <jbk> in it are (for example) mappings 'Offline' and 'Online' whose class keys match the OIDs for the corresponding SNMP traps

[23-Sep-2011 14:41:17] <jbk> if i create an additional mapping without a class key

[23-Sep-2011 14:41:28] <jbk> it will also get applied to any Offline or Online events

[23-Sep-2011 14:42:10] <cluther> jbk: No, definitely not. The first mapping to match an event will be used and no other mappings will be used.

[23-Sep-2011 14:42:32] <cluther> Rather than creating a mapping without a class key you should go to the transform for the /Storage/NetApp/Volume class.

[23-Sep-2011 14:42:59] <cluther> If you're unfamiliar with event class transforms, you can find them by navigating to the event class, clicking the gear menu in the bottom-left and choosing Transform.

[23-Sep-2011 14:43:17] <cluther> This transform code will execute for all events that come into this class or subclasses.

[23-Sep-2011 14:43:24] <jbk> oh.. not where i was expecting it

[23-Sep-2011 14:46:34] <jbk> also, for the python code, can i assume the re module is imported?

[23-Sep-2011 14:47:10] <cluther> jbk: No. You need to import it.

[23-Sep-2011 14:47:30] <jbk> any suggestions for local variable names to avoid clashing?

[23-Sep-2011 14:48:40] <jbk> obviouly in the context the transform code is run, evt is defined

[23-Sep-2011 14:48:52] <cluther> The context will have Python built-ins plus these: evt, device, dev, contToUnits, zdecode, txnCommit, transact, dmd, log

[23-Sep-2011 14:48:55] <cluther> So don't use those.

[23-Sep-2011 14:49:12] <cluther> Well.. you can use them, but don't redefine them.

[23-Sep-2011 14:49:34] <cluther> contToUnits should have been convToUnits.

[23-Sep-2011 14:50:23] <cluther> log is useful to help debug transforms..

[23-Sep-2011 14:50:31] <cluther> log.info('WTF is happening here?')

[23-Sep-2011 14:51:13] <cluther> If you create events manually in the web interface you'll see these logs in event.log. Otherwise you'll find them in zenhub.log.

[23-Sep-2011 14:54:16] <humanix> i have a doubt

[23-Sep-2011 14:54:33] <humanix> can i associate in an arbitrary way a trap comming from an IP to a device

[23-Sep-2011 14:54:39] <humanix> that is not on the OS interfaces

[23-Sep-2011 14:58:16] <cluther> humanix: Only halfway.

[23-Sep-2011 14:59:02] <cluther> You can set the evt.device field in your transform. That'll work to associate the event to the device and you'll see it as such in the global event console and even in the event console for the device.

[23-Sep-2011 14:59:42] <cluther> The caveat is that transforms are applied AFTER the "apply device context" step of event processing. What this means is that the event won't be tagged with all of the groups, systems, locations, production state and priority of the device.

[23-Sep-2011 15:02:54] <humanix> thanks cluther

[23-Sep-2011 15:25:03] <jb> cluther: so, should the step from 4.0.2 -> 4.2ent be painless?

[23-Sep-2011 15:25:06] <jb>

[23-Sep-2011 15:25:49] <rocket> jb: lol we dont know everything changing in 4.2 yet ..

[23-Sep-2011 15:25:59] <jb> ugh

[23-Sep-2011 15:26:09] <jb> well, i

[23-Sep-2011 15:26:16] <jb> well, i'm going to roll 4.0.2.

[23-Sep-2011 15:26:30] <cluther> jb: The big upgrade concern is migrating from MySQL to PostgreSQL.

[23-Sep-2011 15:26:35] <jb> ?!

[23-Sep-2011 15:28:03] <jb> when is that happening?

[23-Sep-2011 15:28:25] <cluther> The strong indication is 4.2.

[23-Sep-2011 15:28:28] <rocket> jb: I believe 4.2

[23-Sep-2011 15:28:55] <jb> that sounds painful.

[23-Sep-2011 15:29:28] <rocket> shouldnt be that bad I dont think.  I believe the new event system is much easier to migrate than the one in 3.X

[23-Sep-2011 15:29:37] <Hackman238> jb: Shouldnt be

[23-Sep-2011 15:29:37] <cluther> Oracle is making using MySQL as painful as it possibly can.

[23-Sep-2011 15:30:02] <Hackman238> cluther: Zends not safe from oracle?

[23-Sep-2011 15:30:06] <jb> ok, so will I be kicking myself in the ass if I deploy 4.0.2?

[23-Sep-2011 15:30:10] <jb> thats the real question

[23-Sep-2011 15:30:16] <Hackman238> cluther: Or does oracle not like refriend mysql? (Zends)

[23-Sep-2011 15:30:21] <rocket> jb: no

[23-Sep-2011 15:30:23] <cluther> jb: No.

[23-Sep-2011 15:30:27] <jb> ok good

[23-Sep-2011 15:30:48] <cluther> Hackman238: zends only exists because of Oracle turning the screws. It's an expensive option.

[23-Sep-2011 15:31:09] <Hackman238> cluther: ah gotcha

[23-Sep-2011 15:31:11] <rocket> even zends itself isnt cheap

[23-Sep-2011 15:31:31] <Hackman238> I thought it would have been cheaper since its rebranded

[23-Sep-2011 15:31:56] <cluther> Hackman238: You're right about that. But "cheaper" means something different when it comes to Oracle.

[23-Sep-2011 15:32:14] <cluther> We're working hard to make sure less of your money ends up in Oracle's pockets.

[23-Sep-2011 15:32:20] <Hackman238> cluther: yeah I know what you mean

[23-Sep-2011 15:32:49] <Hackman238> cluther: Thanks!

[23-Sep-2011 15:34:05] <Hackman238> cluther: Evil folks over at Oracle

[23-Sep-2011 15:34:08] <Hackman238> LOL

[23-Sep-2011 15:34:22] <bender183> pos oracle

[23-Sep-2011 15:34:25] <bender183>

[23-Sep-2011 15:35:14] <SDuensin> Ugh.  Reading technical stuff and code while sick is not a good time.  Nothing makes sense today.

[23-Sep-2011 15:35:55] <Hackman238> SDuensin: Never work sick

[23-Sep-2011 15:36:04] <SDuensin> Hackman238, yea, but I like my job.  :-)

[23-Sep-2011 15:36:11] <Hackman238> SDuensin: You'll take longer to get better and frustrate everyone including yourself

[23-Sep-2011 15:36:26] <Hackman238> SDuensin: Your job doesnt value you if they dont value you and your coworkers health

[23-Sep-2011 15:36:36] <SDuensin> Besides, I needed to be online in case I had to talk to that Hackman238 character.  :-)

[23-Sep-2011 15:36:56] <Hackman238> SDuensin: LOL

[23-Sep-2011 15:37:15] <jbk> i'm assuming setting the zEventSeverity to 'clear' causes it to clear any instances of the event class on the device (w/ the same dedup id)?

[23-Sep-2011 15:37:17] <Hackman238> SDuensin: I just got schooled in new ZenPackPersistance madness

[23-Sep-2011 15:37:24] <SDuensin> I can pull anything I need from WMI now, but I have no idea how to jam it into Zenoss.  Single values aren't hard, but what about something like the software inventory or an event log?

[23-Sep-2011 15:37:41] <Hackman238> SDuensin: Win for support today for awesomeness

[23-Sep-2011 15:38:00] <Hackman238> SDuensin: You want to create events?

[23-Sep-2011 15:38:10] <SDuensin> I don't know what I want to do.  :-)

[23-Sep-2011 15:38:19] <Hackman238> SDuensin: Ah- that complicates it all

[23-Sep-2011 15:38:19] <SDuensin> I want Zenoss to know everything about everything!

[23-Sep-2011 15:38:20] <Hackman238> LOL

[23-Sep-2011 15:39:16] <SDuensin> I wrote a nice little utility I can run on a Windows box via ssh to execute wql queries on it.  It returns the results in a pipe-delimited format like the Linux wmic tool.

[23-Sep-2011 15:39:18] <rocket> Hackman238: schooled? lol

[23-Sep-2011 15:39:30] <SDuensin> Of course, I can change any of it now that it works.

[23-Sep-2011 15:39:54] <Hackman238> rocket: Man I bet no one outside Zenoss has seen that refreshCatalog method LOL

[23-Sep-2011 15:40:42] <cluther> I bet the Plone people are all over it. It's part of ZCatalog.

[23-Sep-2011 15:42:01] <Hackman238> cluther: Very true, but its hard to make that connection to Zenoss sice I dont do much plone work.

[23-Sep-2011 15:42:27] <Hackman238> SDuensin: Very cool

[23-Sep-2011 15:42:36] <SDuensin> Hackman238, for example, I can run the following to get a software inventory:   wql select vendor, description, installdate from win32_product

[23-Sep-2011 15:42:41] <SDuensin> But then what?

[23-Sep-2011 15:43:03] <Hackman238> SDuensin: You'll want to write a parser to break it up in to prodcut entries

[23-Sep-2011 15:43:08] <Hackman238> *product

[23-Sep-2011 15:43:31] <SDuensin> Ok, but where do I put it?  Just write it directly into the DMD?

[23-Sep-2011 15:43:37] <Hackman238> SDuensin: that way you could use the existing products/manufac classes

[23-Sep-2011 15:43:49] <Hackman238> SDuensin: Let me find soe infos for you

[23-Sep-2011 15:43:54] <SDuensin> Thanks.

[23-Sep-2011 15:46:02] <jb> did I see a doc once that had performance best practices?

[23-Sep-2011 15:46:08] <Hackman238> SDuensin: See $ZENHOME/Products/DataCollector/CommandParsers/ for examples

[23-Sep-2011 15:46:19] <SDuensin> Hackman238, thank you.  Looking!

[23-Sep-2011 15:46:31] <Hackman238> jb: Yeah, there is one. Look for zope tuning

[23-Sep-2011 15:46:40] <Hackman238> jb: Something along those lines

[23-Sep-2011 15:46:48] <jb> I was looking for filesystem reccomendations, etc.

[23-Sep-2011 15:47:06] <Hackman238> jb: IO problem?

[23-Sep-2011 15:47:11] <jb> hrm, 8 cores is the minimum reccomended for 4.0.2?

[23-Sep-2011 15:47:15] <jb> oh well, this box only has 4

[23-Sep-2011 15:47:19] <jb> nah, new install.

[23-Sep-2011 15:47:19] <Hackman238> SDuensin: the Unmae_a plugin is pretty easy to understand

[23-Sep-2011 15:47:30] <jb> was thinking of putting mysql on it's own logical volume.

[23-Sep-2011 15:47:42] <Hackman238> jb: You're going to want lots of RAM and very fast disks

[23-Sep-2011 15:47:54] <jb> i have 16GB and a 15KRPM RAID-10 with 4 spindles just for the data

[23-Sep-2011 15:48:01] <SDuensin> Hackman238, was just viewing that one.  Says it returns a 'datamap' - is there a list of those somewhere?

[23-Sep-2011 15:48:20] <Hackman238> jb: Depending on how much you intend to monitor, your disks are likely going to be a bottle neck

[23-Sep-2011 15:48:26] <rocket> jb there are more daemons in 4.0.2 I would recommend you get the additional 4 cores

[23-Sep-2011 15:48:27] <jb> just about 300 devices

[23-Sep-2011 15:48:43] <jb> hm.

[23-Sep-2011 15:48:52] <Hackman238> jb: Ah, then you'll be site long as not 600 processes on each server

[23-Sep-2011 15:49:21] <jb> crap, i thought this box had another CPU in it.

[23-Sep-2011 15:49:24] <rocket> it might be ok on 4 but if it is slow we will tell you to upgrade the box

[23-Sep-2011 15:49:25] <Hackman238> jb: I agree with rocket, you'll have lots of scheduler competition with so few cores

[23-Sep-2011 15:49:48] <rocket> there are ~3 new javaprocesses with the new event system

[23-Sep-2011 15:49:50] <Hackman238> jb: Try not to count hyper threading logical cores as cores...their not and usless for a variety of operations

[23-Sep-2011 15:49:55] <rocket> and one of those spawns workers

[23-Sep-2011 15:49:58] <jb> well yeah.

[23-Sep-2011 15:50:06] <jb> lemme see if I have another R710 laying around.

[23-Sep-2011 15:50:36] <Hackman238> jb: I dont mean to say something you already know. Unfortunately many people count the HT cores as real cores.

[23-Sep-2011 15:50:46] <rocket> however at 8core/8gb of ram zenoss can crunch some data ..

[23-Sep-2011 15:51:01] <jb> got the NOC guys looking for another box now

[23-Sep-2011 15:51:48] <Hackman238> SDuensin: Dont understand

[23-Sep-2011 15:52:41] <SDuensin> Hackman238, well say I write something to parse the results of my wql command...  Where do I store it?  What do I write those values into?

[23-Sep-2011 15:52:42] <Hackman238> SDuensin: when the command parser is selected for a command datasource it will be used to process the command output and map output to datapoints

[23-Sep-2011 15:53:19] * SDuensin is thinking of the software list again, not just a single data point.

[23-Sep-2011 15:53:40] <Hackman238> SDuensin: or right right, by bad

[23-Sep-2011 15:53:54] <Hackman238> SDuensin: thats going to come down to your modeler

[23-Sep-2011 15:53:59] <Hackman238> SDuensin: 1 second

[23-Sep-2011 15:54:56] <Hackman238> SDuensin: $ZENHOME/Products/DataCollector/plugins/zenoss/snmp

[23-Sep-2011 15:55:00] <Hackman238> (for example)

[23-Sep-2011 15:55:17] <Hackman238> SDuensin: contains HRSWInstalledMap.py, a snmp installed software modeller

[23-Sep-2011 15:55:43] <Hackman238> SDuensin: You can write a modeler for ZENHOME/Products/DataCollector/plugins/zenoss/python/ that makes your wmi calls in the smae manner

[23-Sep-2011 15:56:21] <Hackman238> SDuensin: To add it to a zenpack, in your zenpack directory in ./modeler/plugin/python/ create a _init_.py and YourWMIModelerName.py

[23-Sep-2011 15:56:47] <SDuensin> Ok, thanks again man.  I'll do a lot more reading.

[23-Sep-2011 15:57:07] <Hackman238> SDuensin: in your python modeler you could use pythin WMI methods or just os.system `wmic blah`

[23-Sep-2011 15:57:35] <SDuensin> I want mine.  :-)

[23-Sep-2011 15:57:39] <Hackman238> SDuensin: So basically you'll find your YourWMIModelerName.py listed in the modeler plugins list as YourWMIModelerName

[23-Sep-2011 15:57:49] <Hackman238> SDuensin: Good luck!

[23-Sep-2011 15:58:07] <SDuensin> I have a feeling you'll hear from me again before it's over.  :-)

[23-Sep-2011 15:58:20] <Hackman238> SDuensin: NP, I'm here most of the time

[23-Sep-2011 16:03:34] <jb> k, found a box with two E5630's

[23-Sep-2011 16:03:43] <jb> yay.

[23-Sep-2011 16:20:52] <rocket> bleh avoid the os.system calls if possible

[23-Sep-2011 16:21:13] <rocket> its a significant resource hog to have to reinitialize the new process etc

[23-Sep-2011 16:23:28] <SDuensin> I agree, but the difference between my binary and the Python WMI stuff is that mine actually works.  :-P

[23-Sep-2011 16:23:45] <SDuensin> I'm looking at pooling all my SSH nonsense so it doesn't beat the remote boxes to death.

[23-Sep-2011 17:00:54] <Hackman238> Have to run

[23-Sep-2011 17:01:52] <Hackman238> Later all!

[23-Sep-2011 17:18:05] kocolosk is now known as kocolosk|away

[23-Sep-2011 19:00:52] <jbk> is there a zenpack that has event mappings for IF-MIB link down & link up events?

[23-Sep-2011 19:01:12] <jbk> (figure i'd save myself from reinventing the wheel if i can)

[23-Sep-2011 20:05:48] <froztbyte> jbk: they're in by default, I think

[23-Sep-2011 20:06:12] <nonsenso> anyone watching the zenoss/puppetconf talk right now?

[23-Sep-2011 20:06:20] <nonsenso> http://puppetlabs.com/puppetconf-live/

[23-Sep-2011 20:08:15] <jbk> froztbyte: the mib values don't appear to be (in 3.1)

[23-Sep-2011 20:08:34] <jbk> but i found the mapping

[23-Sep-2011 20:08:46] <jbk> so i just changed the event key to the OID

[23-Sep-2011 20:57:40] <bender183> nonsenso, is it still going?

[23-Sep-2011 21:20:51] <nonsenso> bender183: nah it was only for 45m

[23-Sep-2011 21:44:47] <rocket> I would have been there .. but had jury duty .. :/

[23-Sep-2011 22:48:53] <jbk> both the event class key and rule are used to select an event class mapping, correct?

[24-Sep-2011 00:00:01] [disconnected at Sat Sep 24 00:00:01 2011]

[24-Sep-2011 00:00:01] [connected at Sat Sep 24 00:00:01 2011]

[24-Sep-2011 00:00:11] [zenoss-logger (logger bot) has joined #zenoss]

[24-Sep-2011 04:25:25] <holmes.freenode.net> *** Notice -- TS for #zenoss changed from 1316836795 to 1164523381

[24-Sep-2011 09:44:24] <blackflag> Hello all, Im new to zenoss. I installed it on ubuntu server 10 via apt. I only wnat to ping some devices but getting always an event: Msg: Misuse of shell builtins . Whats have I done wrong. Can someone help?

[24-Sep-2011 10:25:11] kocolosk|away is now known as kocolosk

[24-Sep-2011 10:56:07] kocolosk is now known as kocolosk|away

[24-Sep-2011 12:00:01] [disconnected at Sat Sep 24 12:00:01 2011]

[24-Sep-2011 12:00:01] [connected at Sat Sep 24 12:00:01 2011]

[24-Sep-2011 12:00:14] [zenoss-logger (logger bot) has joined #zenoss]

[24-Sep-2011 18:00:16] gvlx_ is now known as gvlx

[24-Sep-2011 20:00:34] <fus10nx> anyone around ?

[25-Sep-2011 00:00:01] [disconnected at Sun Sep 25 00:00:01 2011]

[25-Sep-2011 00:00:01] [connected at Sun Sep 25 00:00:01 2011]

[25-Sep-2011 00:00:12] [zenoss-logger (logger bot) has joined #zenoss]

[25-Sep-2011 00:00:13] <gibson.freenode.net> [freenode-info] channel flooding and no channel staff around to help? Please check with freenode support: http://freenode.net/faq.shtml#gettinghelp

[25-Sep-2011 12:00:01] [disconnected at Sun Sep 25 12:00:01 2011]

[25-Sep-2011 12:00:01] [connected at Sun Sep 25 12:00:01 2011]

[25-Sep-2011 12:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[26-Sep-2011 00:00:01] [disconnected at Mon Sep 26 00:00:01 2011]

[26-Sep-2011 00:00:01] [connected at Mon Sep 26 00:00:01 2011]

[26-Sep-2011 00:00:11] [zenoss-logger (logger bot) has joined #zenoss]

[26-Sep-2011 00:32:01] <Diddi> hi! I'm having trouble changing IP on one of my devices. Checking the zope logs it says the IP is already in use, but I just can't find out what device is using that ip (that is, no one should)

[26-Sep-2011 00:40:36] <Diddi> I may have used that IP on one device, and later on changed it to something else. Is there a way to see if that's my  problem?

[26-Sep-2011 05:14:18] <tsener> hmm

[26-Sep-2011 05:14:57] <tsener> How do I get graph for a certain period, for a given start and end date ? is this possible, if not through the zenoss interface, in some other way ?

[26-Sep-2011 07:38:13] <tsener> hmm

[26-Sep-2011 07:38:29] <tsener> where in zenoss 3.2 can I set the mysql DB listening port for the event DB ?

[26-Sep-2011 08:28:34] <ion-fan> i want to participate in the zenoss forums but after the registration confirmation i get "your account is waiting to be approved" message

[26-Sep-2011 08:28:48] <ion-fan> but i dont get a second confirmation mail

[26-Sep-2011 08:28:55] <ion-fan> what do i need to do?

[26-Sep-2011 09:40:06] <JohnnyNOC> Hackman238

[26-Sep-2011 09:40:08] <JohnnyNOC> i LOL'd

[26-Sep-2011 09:40:13] <JohnnyNOC> "Shane Scott you are banned from my posts from here on. You are not helpful and only insult. Thank you."

[26-Sep-2011 09:44:43] <dhopp> JohnnyNOC: where was that at?

[26-Sep-2011 09:45:05] <jmp242> what?

[26-Sep-2011 09:45:17] <JohnnyNOC> forums

[26-Sep-2011 09:45:20] <JohnnyNOC> shane was trying to help some guy

[26-Sep-2011 09:45:59] <JohnnyNOC> "please HELP zenoss install will not start!!!"

[26-Sep-2011 09:46:00] <JohnnyNOC> topic

[26-Sep-2011 09:46:28] <jmp242> linky?

[26-Sep-2011 09:47:48] <JohnnyNOC> message/61482#61482

[26-Sep-2011 09:47:53] <JohnnyNOC> (i haven't read the thread myself)

[26-Sep-2011 09:48:05] <JohnnyNOC> just laughed when i saw that, shane is always so helpful

[26-Sep-2011 09:50:00] <jmp242> It's the infamous TitoOrtega

[26-Sep-2011 09:50:09] <jmp242> that guy is a troll as far as I can tell

[26-Sep-2011 09:50:34] <JohnnyNOC> infamous?

[26-Sep-2011 09:50:34] <JohnnyNOC> hahha

[26-Sep-2011 09:50:40] <JohnnyNOC> why so

[26-Sep-2011 09:50:57] <jmp242> Oh, you should look up all his threads

[26-Sep-2011 09:51:16] <jmp242> he's like an epic troll, the sort of which I haven't seen in years on the net

[26-Sep-2011 09:51:29] <jmp242> They usually get banned on moderated forums

[26-Sep-2011 09:52:36] <JohnnyNOC> wish i had some more time

[26-Sep-2011 09:54:17] <ion-fan> can you help me access the forums?

[26-Sep-2011 09:54:43] <ion-fan> or do i only have to wait

[26-Sep-2011 09:58:24] <humanix> hi there

[26-Sep-2011 10:02:18] <Hackman238> Check out the thread now

[26-Sep-2011 10:05:25] <davetoo> donde?

[26-Sep-2011 10:05:49] <davetoo> (I figure, if you're where I know you are, you know what that means

[26-Sep-2011 10:06:45] <davetoo> (or do you work remotely in some other state?)

[26-Sep-2011 10:11:09] <Hackman238> davetoo: Donde?

[26-Sep-2011 10:11:37] <Hackman238> ion-fan: Have you registered?

[26-Sep-2011 10:12:21] <ion-fan> yes i have, i also got a registration confirmation mail

[26-Sep-2011 10:12:51] <Hackman238> ion-fan: Gotcha. And you still cannot login?

[26-Sep-2011 10:13:00] <ion-fan> yes..

[26-Sep-2011 10:13:13] <ion-fan> you know the solution?

[26-Sep-2011 10:13:13] <Hackman238> ion-fan: Thats odd. What happens?

[26-Sep-2011 10:13:30] <ion-fan> i get "your account is waiting to be approved" message

[26-Sep-2011 10:13:45] <Hackman238> ion-fan: Ah, must be new forum security.

[26-Sep-2011 10:13:50] <Hackman238> ion-fan: Email Nick Yeates <nyeates@zenoss.com>

[26-Sep-2011 10:14:16] <Hackman238> ion-fan: And request your account be enabled. It shuold happen automatically.

[26-Sep-2011 10:14:17] <ion-fan> is there a manual confirmation to get into the forum?

[26-Sep-2011 10:14:25] <davetoo> =="where"

[26-Sep-2011 10:14:42] <Hackman238> ion-fan: It used to be automatic. It might be manual now, Nick has made lots of adjustments lately.

[26-Sep-2011 10:14:48] <Hackman238> davetoo: Ah

[26-Sep-2011 10:14:57] <davetoo> there's been hella-spam.

[26-Sep-2011 10:14:58] <ion-fan> ah ok... thx

[26-Sep-2011 10:15:40] <Hackman238> ion-fan: NP. In the mean time we're happy to help here

[26-Sep-2011 10:16:13] <Hackman238> davetoo: I know about as much Spanish as I do Japanese

[26-Sep-2011 10:17:42] <davetoo> heh... I was on the phone the other day with one of our (Canadian) engineers, and he started chewing out one of his kids in what sounded like perfect japanese

[26-Sep-2011 10:17:51] <davetoo> you just never know

[26-Sep-2011 10:19:24] <davetoo> I still want to know who is the Zope Community Architecture    Or the Zenoss Component Alliance.

[26-Sep-2011 10:19:26] <Hackman238> davetoo: LOL!

[26-Sep-2011 10:20:17] <Hackman238> davetoo: We didn't think of the confusion with zope componenet arch.

[26-Sep-2011 10:20:40] <davetoo> it's just funny because I'm neck-deep in zope component stuff lately

[26-Sep-2011 10:21:44] <davetoo> bbl

[26-Sep-2011 10:28:06] <dpetzel> any idea on why on a 3.0.3 --> 3.1 upgrade zenrender is now longer showing up on my list of daemons, or when I do a zenoss status? In turn its not being start along with the rest

[26-Sep-2011 10:41:55] <dhopp> dpetzel: I think the way the graphs render is changed.  Do you have more than on collector?

[26-Sep-2011 10:42:05] <dhopp> changed by default I mean

[26-Sep-2011 10:42:08] <dpetzel> yeah, 1 'master' two remote collectors

[26-Sep-2011 10:43:32] <dhopp> dpetzel: you can add zenrender to the daemons.txt and then in Advanced->collectors->(choose collector)->Edit-> Render URL->http://collector:8091

[26-Sep-2011 10:45:43] <dpetzel> I can use the colloector:8091 approach, as my GUI clients have no access to the collector, I had to configure it http://MASTER:8090/collector

[26-Sep-2011 10:45:50] <dpetzel> so it routes through the master

[26-Sep-2011 10:45:57] <dpetzel> I cant* use is what I meant

[26-Sep-2011 10:46:08] <dhopp> oh..in that case I don't think you need zenrender

[26-Sep-2011 10:47:31] <dpetzel> hmm..OK.. So Maybe I jumped the gun, I assumed ZenRender was my problem since my graphs were not rendering for devices on the collectors

[26-Sep-2011 10:49:34] <jmp242> Hackman238: hey, can anyone take a look at these two forum threads and give an answer a go? message/61485#61485

[26-Sep-2011 10:49:42] <jmp242> message/61481#61481

[26-Sep-2011 10:50:41] <dpetzel> I'm getting the following: ProtocolError: <ProtocolError for masterserver:8090/collector: 500 Internal Server Error>

[26-Sep-2011 10:51:47] <dhopp> dpetzel:  hrm…you can try to start zenrender on the collector manually to see if that does resolve it (zenrender start as the zenoss user)

[26-Sep-2011 10:52:43] <dpetzel> yeah in the process of doing that, they in the process lost the config as well, so fixing that( I believe the config was lost when I did an update remote collectors, before I noticed zenrender not running on the master)

[26-Sep-2011 10:53:18] <dhopp> dpetzel:  are you an enterprise customer?

[26-Sep-2011 10:53:24] <dpetzel> negative

[26-Sep-2011 10:53:54] <dhopp> so are you using Egor's zenpack?

[26-Sep-2011 10:54:01] <dpetzel> dhopp: affirmative

[26-Sep-2011 10:54:35] <dhopp> you might be able to set the render url to /zport/RenderServer…

[26-Sep-2011 10:54:58] <dhopp> Do you have the latest version of his zenpack?  I know a revision or 2 ago he had a problem that didn't start zenrender

[26-Sep-2011 10:55:23] <dpetzel> dhopp: just checking that now actually, appears to be the latest 1.7, trying to re-install it now to see if helps

[26-Sep-2011 10:57:54] <dpetzel> after reinstalling the pack, zenrender is once again in the list of daemons, updating remotes now

[26-Sep-2011 10:59:18] <crazed> is there a way to remove a patch applied with zenpatch?

[26-Sep-2011 11:00:23] <dhopp> crazed: I think you can run zenpatch again and it will ask you if you want to remove it

[26-Sep-2011 11:01:24] <crazed> oh really?

[26-Sep-2011 11:01:25] <crazed> cool

[26-Sep-2011 11:01:28] <crazed> just wondering

[26-Sep-2011 11:01:50] <crazed> so ever since upgrading to 3.2.0 i get a lot of device timeouts from snmp

[26-Sep-2011 11:01:55] <jrh0090> there should also be a .orig file next to the one that was patched if you want to see what changed

[26-Sep-2011 11:02:01] <crazed> did the snmp timeout get lowered?

[26-Sep-2011 11:03:06] <crazed> zSnmpTimeout is set to 20 though hm

[26-Sep-2011 11:04:43] <dpetzel> dhopp: after reinstalling the pack an udpating remotes, graphs are rendering again so that is good. However zoom appears to have no result anymore (which I think I recall some posts about that so searching through them now)

[26-Sep-2011 11:06:29] <crazed> i have a general question, what version of snmp should be used for the best performance?

[26-Sep-2011 11:06:33] <crazed> i've got 600 devices

[26-Sep-2011 11:06:38] <crazed> and currently it's set to v1

[26-Sep-2011 11:06:46] <dhopp> v2c at least

[26-Sep-2011 11:07:04] <dhopp> v2c is way better for performance..it's less load on the monitored device and zenoss

[26-Sep-2011 11:07:06] <pipeline> you probably want v2c just to support gigabit ethernet

[26-Sep-2011 11:07:14] <pipeline> v1 counters roll over too fast to get accurate counts

[26-Sep-2011 11:07:32] <crazed> hm

[26-Sep-2011 11:07:34] <crazed> thanks pipeline

[26-Sep-2011 11:07:56] <pipeline> if security is an issue at all use the newest v3 implementation you can lay hands on

[26-Sep-2011 11:08:10] <pipeline> (e.g. v3 is secure enough to use over the open internet but only if you have newest versinos of net-snmp on both sides)

[26-Sep-2011 11:08:13] <crazed> security isn't really an issue

[26-Sep-2011 11:08:22] <pipeline> there you have it then

[26-Sep-2011 11:08:27] <crazed> awesome

[26-Sep-2011 11:08:41] <crazed> i'll switch over to v2c on a few devices and see if anything starts breaking

[26-Sep-2011 11:08:47] <crazed> before making a global change

[26-Sep-2011 11:09:08] <pipeline> so I am trying to use a python expression for the "maximum value" of a threshold

[26-Sep-2011 11:09:16] <pipeline> and it just doesn't seem to be working

[26-Sep-2011 11:09:20] <pipeline> how do I troubleshoot this ?

[26-Sep-2011 11:16:01] <Hackman238> pipeline: Whats the expression?

[26-Sep-2011 11:16:34] <Hackman238> pipeline: Often the problem is divide by zero, bad type or empty variables.

[26-Sep-2011 11:17:28] <Hackman238> pipeline: If you have a variable you want to use, ex speed, use (here.speed or 100000). This falls back to the value right of or of here.speed is invalid

[26-Sep-2011 11:18:11] <pipeline> oh that's a great tip

[26-Sep-2011 11:19:26] <Hackman238> pipeline: Let me know if you need any help!

[26-Sep-2011 11:19:51] <pipeline> and you more or less divined what I'm attempting to do

[26-Sep-2011 11:19:59] <pipeline> I have a single data source (HeapMemoryUsage) with two data points (used and committed)

[26-Sep-2011 11:20:09] <pipeline> I'm trying to threshold when "used" gets too close to "committed"

[26-Sep-2011 11:21:08] <Hackman238> pipeline: Ah gotcha

[26-Sep-2011 11:22:13] <pipeline> right now I have a threshold configured on "committed," with a max expression of: (here.used or 10000)

[26-Sep-2011 11:22:19] <pipeline> just to test thresholding

[26-Sep-2011 11:22:37] <pipeline> the "committed" variable is *always* higher than "used", so an event should be created

[26-Sep-2011 11:22:45] <pipeline> trying to figure out what the heck is going on

[26-Sep-2011 11:24:52] <pipeline> when I run "zenjmx -v 10 run --device ...", there are no zen.threshold lines at all for my non-working threshold

[26-Sep-2011 11:25:01] <pipeline> so I infer something must be garbage in my expression itself?

[26-Sep-2011 11:28:30] <ion-fan> i there a way to search globally in the interface descriptions like here(message/56987) but in the gui?

[26-Sep-2011 11:29:19] <crazed> hm so i switched the zSnmpVer to v2c at /, but seems existing hosts still have snmp version: v1 set, though i see the zSnmpVer set to v2c, will these update themselves when they get remodeled?

[26-Sep-2011 11:29:34] <pipeline> crazed: I think you hit the nail on the head

[26-Sep-2011 11:32:14] <Hackman238> ion-fan: My updated device search pack can do that

[26-Sep-2011 11:32:37] <ion-fan> can you give me an url?

[26-Sep-2011 11:33:17] <pipeline> Hackman238: is what I'm trying to do even reasonable? (i.e. use a second data point in the first data point's threshold)

[26-Sep-2011 11:33:41] <Hackman238> ion-fan: http://shanewilliamscott.com/Code/devicepropertysearch-zenpack.html

[26-Sep-2011 11:34:10] <dhopp> pipeline:  there is a zenpack that does that I believe…I don't think zenoss does it by default

[26-Sep-2011 11:34:11] <Hackman238> ion-fan: You'll need to reconfigure it for device.os.interfaces.interface.description

[26-Sep-2011 11:34:35] <pipeline> dhopp: so what can I do with python expressions in the "Maximum value" field... what's permissible ?

[26-Sep-2011 11:34:41] <Hackman238> pipeline: Is used a rrd value/datapoint or an attribute?

[26-Sep-2011 11:34:44] <ion-fan> in this file? Edit '%ZENHOME/ZenPacks/ZenPacks.ShaneScott.DeviceSearch-1.0.0-

[26-Sep-2011 11:34:44] <ion-fan> py2.6.egg/ZenPacks/ShaneScott/DeviceSearch/__init__.py'

[26-Sep-2011 11:34:53] <Hackman238> ion-fan: Yes

[26-Sep-2011 11:35:29] <pipeline> Hackman238: it's a data point from the JMX collector.  I don't know what an attribute would be in the context of a data source

[26-Sep-2011 11:38:20] <Hackman238> here.getRRDValues('used') is what you need, I think

[26-Sep-2011 11:38:50] <Hackman238> pipeline: It could impact performance if used on a large scale though

[26-Sep-2011 11:39:12] <Hackman238> pipeline: (here.getRRDValues('used') or 10000)

[26-Sep-2011 11:39:21] <Hackman238> pipeline: Just incase used == NaN

[26-Sep-2011 11:39:46] <pipeline> yeah

[26-Sep-2011 11:39:51] <pipeline> that seems plausible

[26-Sep-2011 11:40:27] <ion-fan> Hackman238: when i execute the zenpack --install=... - command i got the error "WARNING zen.Events: Error processing transform/mapping on Event Class /

[26-Sep-2011 11:40:28] <ion-fan> Problem on line 2: AttributeError: 'NoneType' object has no attribute 'prepId'

[26-Sep-2011 11:40:28] <ion-fan> fs_id = device.prepId(evt.component)"

[26-Sep-2011 11:40:57] <ion-fan> i have already certain transforms in my zenoss instance..

[26-Sep-2011 11:43:14] <enkrypt> quite often I get this scary event on my zenoss hub: MySQL Error: (1129, "Host 'zenoss2.cegeka.be' is blocked because of many connection errors; unblock with 'mysqladmin flush-hosts'")

[26-Sep-2011 11:43:25] <pipeline> Hackman238: thanks, man.  It's still not working, but now it prints useful things to the log from zen.thresholds, so I must be barking up the right tree.

[26-Sep-2011 11:47:28] <enkrypt> doing a mysqladmin flush-host solves the problem, but I am still a bit worried about this, it seems my collectors are sometimes causing too many mysql connection errors, which causes them to get blacklisted

[26-Sep-2011 11:50:34] <crazed> zenoss supports the bulk feature of v2c without me changing any other setting, correct?

[26-Sep-2011 11:51:12] jellis is now known as Poised

[26-Sep-2011 12:00:01] [disconnected at Mon Sep 26 12:00:01 2011]

[26-Sep-2011 12:00:02] [connected at Mon Sep 26 12:00:02 2011]

[26-Sep-2011 12:00:15] [zenoss-logger (logger bot) has joined #zenoss]

[26-Sep-2011 12:00:16] <sendak.freenode.net> [freenode-info] help freenode weed out clonebots -- please register your IRC nick and auto-identify: http://freenode.net/faq.shtml#nicksetup

[26-Sep-2011 13:01:02] <Hackman238> pipeline: Good luck

[26-Sep-2011 13:01:38] <Hackman238> crazed: yes

[26-Sep-2011 13:10:01] <dhopp> is it possible to use ${dev/id} as part of a zlink?  so the zlink will actually be http://url/mydevice assuming 'mydevice' is the device's id?

[26-Sep-2011 13:10:16] <dhopp> I'm thinking no..but maybe there is some magic I'm not thinking about

[26-Sep-2011 13:11:19] <jbk> is there a way to create an event based on the frequency of another event?

[26-Sep-2011 13:11:35] <jbk> for example, i've got a port right now that seems to be flapping (going down and up)

[26-Sep-2011 13:11:43] <jbk> so i get the link down event

[26-Sep-2011 13:11:47] <jbk> then the link up event clears it

[26-Sep-2011 13:39:25] <dhopp> nevermind it is possible :-)

[26-Sep-2011 13:39:48] <dhopp> <a href="http://url/${here/id}>My Cool Device</a> works...

[26-Sep-2011 13:39:58] kocolosk|away is now known as kocolosk

[26-Sep-2011 14:47:26] <jmp242> afternoon

[26-Sep-2011 14:53:43] <Sashness> afternoon

[26-Sep-2011 15:23:44] <jmp242> Anyone know of windows RPC issues that would have WMI queries fail across several domains per this post? message/61543#61543

[26-Sep-2011 15:23:50] <jmp242> cgibbons ?? lol

[26-Sep-2011 15:41:17] <jbk> what would be the easiest way to set the severity of an alert based on it's location?

[26-Sep-2011 15:41:31] <jbk> or set the action

[26-Sep-2011 15:41:34] <jmp242> probably an event transform

[26-Sep-2011 15:42:51] <jbk> how would one get the location of the device from the event?

[26-Sep-2011 15:42:56] <jbk> evt.device.???

[26-Sep-2011 15:43:21] <jmp242> Look at an event

[26-Sep-2011 15:43:29] <jmp242> see if there's a .location or some such

[26-Sep-2011 15:43:31] <jmp242> but I doubt it

[26-Sep-2011 15:44:00] <jmp242> probably something like the get syntax, but I'm not a coder

[26-Sep-2011 15:44:08] <jmp242> Jane_Curry: ping

[26-Sep-2011 15:44:13] <jmp242> Hackman238: ping

[26-Sep-2011 15:44:20] <jmp242> one of them might be able to answer

[26-Sep-2011 15:44:25] <jmp242> if they're here

[26-Sep-2011 15:44:27] <jmp242> and not afk

[26-Sep-2011 15:57:25] <JohnnyNOC> docs/DOC-4821

[26-Sep-2011 15:57:36] <JohnnyNOC> jbk:  read that ---^

[26-Sep-2011 15:57:52] <JohnnyNOC> will show you how to get location, and over event attributes

[26-Sep-2011 16:00:02] <JohnnyNOC> jbk:  evt.Location

[26-Sep-2011 16:00:37] <jbk> ahh ok..

[26-Sep-2011 16:01:32] <jbk> any ideas how to spot an event that is frequently created + cleared (like repeated link down & link up notifications?)

[26-Sep-2011 16:02:42] <JohnnyNOC> create an email alert for it?

[26-Sep-2011 16:02:55] <JohnnyNOC> nm

[26-Sep-2011 16:02:59] <jbk> that would be a bit obnoxious

[26-Sep-2011 16:03:02] <JohnnyNOC> heh

[26-Sep-2011 16:03:28] <JohnnyNOC> at least that way you could see the alerts/clears in zenactions.log

[26-Sep-2011 16:03:34] <JohnnyNOC> though i'm sure there's a much more intelligent way

[26-Sep-2011 16:03:35] <JohnnyNOC>

[26-Sep-2011 16:03:58] <JohnnyNOC> otherwise, i guess you have to peek into the events db

[26-Sep-2011 16:04:10] <jbk> as i've added our switches into the mix, i've foudn a few ports whose links are constantly going up and down (bad cable)

[26-Sep-2011 16:04:22] <jbk> the linnk up notification clears the link down one

[26-Sep-2011 16:05:03] <jbk> but if it's a fairly constant number of them after 20-30s, then it's probably something worth noting

[26-Sep-2011 16:05:50] <JohnnyNOC> you could write something to check for those events in mysql with a certain count or difference in time for example

[26-Sep-2011 16:05:54] <JohnnyNOC> but again, i think there must be a better way

[26-Sep-2011 18:26:39] kocolosk is now known as kocolosk|away

[26-Sep-2011 19:04:41] <robo> hello: i installed the apache monitor zenpack but it doesn't appear to work. I have server-status working, and if i do a wget to server-status from the zenoss server it works as expected

[26-Sep-2011 19:05:16] <robo> but in the event log i see Server replied: 403 Forbidden to action GET /server-status?auto

[26-Sep-2011 19:05:19] <robo> so i don't get it

[26-Sep-2011 19:06:40] <robo> hmm

[26-Sep-2011 19:06:47] <robo> i wonder if it's trying to use ssl or something

[26-Sep-2011 19:53:04] kocolosk|away is now known as kocolosk

[26-Sep-2011 21:22:15] <robo> i don't get it. I can do a wget from the machine that runs zenoss to server-status but zenoss reports a 403 in events for the apache monitor script

[26-Sep-2011 21:22:45] <robo> any ideas why wget works but the apache monitor fails?

[26-Sep-2011 21:22:54] <robo> is there a command line script i can use to test this?

[26-Sep-2011 21:30:35] <robo> okay, so ./check_apache.py returns 403 forbidden, but wget works fine

[26-Sep-2011 21:58:37] <dhopp> robo: are you doing the wget from the zenoss machine?

[26-Sep-2011 21:58:50] <robo> dhopp, yup

[26-Sep-2011 21:58:55] <robo> i think it's an apache config problem

[26-Sep-2011 21:59:07] <robo> because if i telnet i get a 403. I do a wget and it works. Very very odd

[26-Sep-2011 22:00:18] <dhopp> when you use wget what does your command look like?  is it wget http://dnsname/server-status?auto or http://ipaddress/server-status?auto ?

[26-Sep-2011 22:03:41] <robo> dnsname

[26-Sep-2011 22:03:57] <dhopp> zenoss is using ipaddress by default

[26-Sep-2011 22:04:02] <robo> oooooooooooo

[26-Sep-2011 22:04:06] <dhopp> can you try your wget using the IP address and verify you get a 403?

[26-Sep-2011 22:04:44] <dhopp> the apache monitor has ${dev/manageIP} as the host it requests

[26-Sep-2011 22:04:48] <dhopp> by default

[26-Sep-2011 22:05:49] <robo> it ends up being the same deal. wget on the IP works, web-browser works, zenoss gets a 403

[26-Sep-2011 22:05:55] <robo> and i get a 403 if i use telnet

[26-Sep-2011 22:06:07] <robo> so weird

[26-Sep-2011 22:06:20] <dhopp> anything in the apache logs?

[26-Sep-2011 22:06:49] <robo> not that i notice

[26-Sep-2011 22:06:58] <robo> i wonder if some re-write rule is hit or something

[26-Sep-2011 22:07:20] <dhopp> if that was the case the wget should trigger it as well

[26-Sep-2011 22:07:37] <dhopp> are you using http or https?

[26-Sep-2011 22:16:10] <robo> http

[26-Sep-2011 22:21:06] <dhopp> something is odd.. a 403 implies that there is some kind of acl (most likely under Location /server-status in apache) but that should cause your wget to get a 403 as well assuming you are doing the wget from the same zenoss machine

[26-Sep-2011 22:24:09] <robo> dhopp, it was a weird rewrite rule

[26-Sep-2011 22:24:21] <robo> if no user-agent was passed or no referrer we were 403'ing

[26-Sep-2011 22:58:19] <dhopp> ah

[27-Sep-2011 00:00:01] [disconnected at Tue Sep 27 00:00:01 2011]

[27-Sep-2011 00:00:02] [connected at Tue Sep 27 00:00:02 2011]

[27-Sep-2011 00:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[27-Sep-2011 01:09:33] kocolosk is now known as kocolosk|away

[27-Sep-2011 02:39:14] Poised is now known as jellis

[27-Sep-2011 07:49:30] <ion-fan> anyone a clue? especially hackman238 ? message/61547#61547

[27-Sep-2011 08:00:31] <enkrypt> I get this error often when I apply the Mysql monitoring template to my hub: MySQL Error: (1129, "Host 'zenoss2.cegeka.be' is blocked because of many connection errors; unblock with 'mysqladmin flush-hosts'")

[27-Sep-2011 08:03:55] <tsener> check  show full processlist in the  monitored mysql

[27-Sep-2011 08:04:09] <tsener> and if you have mtop installed there (top for mysql) could give you a clue

[27-Sep-2011 08:04:40] <tsener> probably zenoss tries too many queries  and they do not expire, you might have to tune the my.cnf

[27-Sep-2011 08:22:19] <enkrypt> tsener: show full processlist; shows mostly around 10 sleeping processes under the zenoss account

[27-Sep-2011 08:33:59] <enkrypt> tsener: I did the best I could for my.cnf using the Mysqltuner script and the Zenoss guides

[27-Sep-2011 08:35:26] <enkrypt> mysqltuner warns me about this:  Connections aborted: 83%

[27-Sep-2011 08:35:34] <enkrypt> "Your applications are not closing MySQL connections properly"

[27-Sep-2011 08:36:44] <tsener> the problem might be  with the mysql zenpack or whatever you are using  to connect to mysql

[27-Sep-2011 08:37:24] <tsener> I have  had such problems with php scripts, there the problem was a flag for the mysql_connect() for persistant connections

[27-Sep-2011 08:38:39] <Hackman238> Morning all!

[27-Sep-2011 08:39:13] <jmp242> morning Hackman238

[27-Sep-2011 08:39:29] <jmp242> how's it going

[27-Sep-2011 08:39:44] <enkrypt> tsener: I'm using Zenoss Enterprise 3.2.0, with the 3 zenpatches applied. the version of my MySQLMonitor zenpack is reported as 2.1.3

[27-Sep-2011 08:40:15] <Hackman238> jmp242: Busy busy! Was sick yesterday from insanely hot chicken wings.

[27-Sep-2011 08:40:18] <Hackman238> LOL

[27-Sep-2011 08:40:27] <jmp242> oh man

[27-Sep-2011 08:40:44] <jmp242> I'm finally better too after a second root canal on the same tooth

[27-Sep-2011 08:41:28] <Hackman238> jmp242: Oh damn. Sorry to hear you had to have that done.

[27-Sep-2011 08:41:41] <jmp242> yea, the most painful thing is the cost

[27-Sep-2011 08:41:48] <jmp242> thank goodness for dental insurance

[27-Sep-2011 08:42:08] <Hackman238> jmp242: It was my fault, I choose to keep eating those wings...was trying to beat the challenge LOL

[27-Sep-2011 08:42:19] <Hackman238> jmp242: I can imagine.

[27-Sep-2011 08:42:53] <jmp242> While I've got you

[27-Sep-2011 08:43:10] <jmp242> do you know why Zenoss 3.1 would seem to not really re-model devices automatically?

[27-Sep-2011 08:43:13] <Hackman238> jmp242: I asked what type of peppers they sue in the sauce...they tell me 'we dont use any peppers, jsut a vial of capsasain oil'

[27-Sep-2011 08:43:23] <jmp242> Specifically

[27-Sep-2011 08:43:45] <jmp242> when some of our linux boxes change their filesystem size zenoss doesn't notice until I manually model them

[27-Sep-2011 08:43:45] <Hackman238> jmp242: Yep, zenmodeler bas become the living dead. Its alive and heartbeating but its really dead and needing a zenhub reload then a zenmodeler reload

[27-Sep-2011 08:44:03] <jmp242> so I get the fun -50 GB free space

[27-Sep-2011 08:44:04] <jmp242> lo

[27-Sep-2011 08:44:06] <jmp242> l

[27-Sep-2011 08:44:10] <jmp242> Oh great

[27-Sep-2011 08:44:11] <jmp242> ok

[27-Sep-2011 08:44:43] <Hackman238> jmp242: We have this problem here in v2.5.2 and virtually every client of mine has the problem running v2.5.2 or v3

[27-Sep-2011 08:44:53] <jmp242> If I do that - will it start a remodel of all systems

[27-Sep-2011 08:44:59] <jmp242> or do I need to also do a zenmodeler run

[27-Sep-2011 08:45:24] <Hackman238> jmp242: Best solution is a baby sitter script to keep an eye on zenmodeler- if the log stops growing at a minimum expected rate, kill it and reload it.

[27-Sep-2011 08:45:47] <Hackman238> jmp242: You'll need to jump on each collector and zenmodeler stop/start

[27-Sep-2011 08:45:47] <jmp242> ahh. Hey, do you have one we can release to the community?

[27-Sep-2011 08:46:08] <jmp242> if this is a common problem I'd love to put something on zcaportal (and save me time in re-creating the wheel) lol

[27-Sep-2011 08:46:24] <Hackman238> jmp242: Sure

[27-Sep-2011 08:46:26] <jmp242> I'm simple

[27-Sep-2011 08:46:32] <jmp242> only one zenoss box and one local collector

[27-Sep-2011 08:46:39] <jmp242> it has lots of grunt for ~150 devices

[27-Sep-2011 08:46:50] <Hackman238> jmp242: A simpliler solution is to kill and reload zenmodeler at 00:00 everyday via cron job

[27-Sep-2011 08:47:04] <jmp242> don't I also have to do zenhub

[27-Sep-2011 08:47:06] <jmp242> or no

[27-Sep-2011 08:47:06] <Hackman238> jmp242: Gotcha

[27-Sep-2011 08:47:26] <jmp242> The cron solution is actually fine with me

[27-Sep-2011 08:48:06] <jmp242> to clarify

[27-Sep-2011 08:48:10] <Hackman238> jmp242: Thats probably the better solution. The babysitter method is better when a zenmodeler restart creates crippling waves of zenhub madness in v2.5.2 scaled to epic sizes

[27-Sep-2011 08:48:13] <jmp242> when you say reload, do you mean restart it

[27-Sep-2011 08:48:23] <Hackman238> jmp242: yep

[27-Sep-2011 08:48:27] <jmp242> ok

[27-Sep-2011 08:48:58] <jmp242> I'll need to write this up then on the new FAQ

[27-Sep-2011 08:49:06] <Hackman238> jmp242: Alrighty

[27-Sep-2011 08:49:12] <Hackman238> jmp242: Thanks man

[27-Sep-2011 08:49:57] <jmp242> Thank you

[27-Sep-2011 08:50:08] <jmp242> it would have taken me a bunch of log looking to figure that zenmodeler was a zombie

[27-Sep-2011 08:50:45] <jmp242> Hey, do you want to see some forum posts? There is one on your zenpack

[27-Sep-2011 08:50:58] <Hackman238> jmp242: It could be other things, but its probably just a dead zenmodeler

[27-Sep-2011 08:51:04] <Hackman238> jmp242: I havent- been sick

[27-Sep-2011 08:51:16] <Hackman238> jmp242: except tito's

[27-Sep-2011 08:51:20] <jmp242> I know you've been sick

[27-Sep-2011 08:51:21] <Hackman238> LOL

[27-Sep-2011 08:51:26] <jmp242> I meant, can I send some up here

[27-Sep-2011 08:51:34] <jmp242> message/61547#61547

[27-Sep-2011 08:51:35] <Hackman238> jmp242: oh sure

[27-Sep-2011 08:51:38] <jmp242> Tito is epic troll

[27-Sep-2011 08:51:48] <jmp242> I haven't seen the like in years on moderated forums

[27-Sep-2011 08:52:21] <jmp242> This one is fun actually - I wish cgibbons was still around - he got the weirdness of MS Internal RPC issues message/61543#61543

[27-Sep-2011 08:53:49] <jmp242> This one other IRC people could also jump on if they have trap experience, but traps just seem flaky

[27-Sep-2011 08:53:50] <jmp242> message/61496#61496

[27-Sep-2011 08:54:39] <jmp242> And this one I think I saw something similar, but I don't know that it's relevant, recently; message/61481#61481

[27-Sep-2011 08:57:42] <Hackman238> jmp242: Thanks man. I'll read through those soon

[27-Sep-2011 09:01:45] <enkrypt> what do I do about these kinds of warnings again?: Error reading value for "StackSub-St1-1" on 10.250.0.17 (oid .1.3.6.1.2.1.2.2.1.20.5180 is bad)

[27-Sep-2011 09:01:53] <enkrypt> these show up on certain Cisco switches

[27-Sep-2011 09:02:00] <enkrypt> monitoring via SNMPv2c

[27-Sep-2011 09:02:56] <Hackman238> enkrypt: Add  .0 to the end of the value being polled

[27-Sep-2011 09:03:06] <humanix> hi there

[27-Sep-2011 09:03:17] <Hackman238> enkrypt: In the case of a stack stub, it may not support the oid being polled

[27-Sep-2011 09:03:21] <Hackman238> humanix: Hey

[27-Sep-2011 09:03:26] <humanix> hi Hackman238

[27-Sep-2011 09:03:27] <humanix>

[27-Sep-2011 09:03:35] <jb> hrm, how hard would it be to have multiple templates for email alerts?  the standard alert is hard to read on a phone as a SMS.

[27-Sep-2011 09:03:43] <humanix> hackman do you know how to generate a FS exclution

[27-Sep-2011 09:03:47] <jb> i'd like to create a simpler alert just for SMS purposes.

[27-Sep-2011 09:03:58] <humanix> cause proc fs is alerting me xD

[27-Sep-2011 09:04:39] <Hackman238> jb: Just create a copy of the alerting rule that only sends pages and adjust the message there

[27-Sep-2011 09:05:04] <Hackman238> humanix: You'll want to add it to a z property for exclusion

[27-Sep-2011 09:05:06] <Hackman238> humanix: 1 sec

[27-Sep-2011 09:05:55] <Hackman238> humanix: add proc to 'zFileSystemMapIgnoreNames'

[27-Sep-2011 09:06:18] <Hackman238> humanix: Then remodel. If proc doesnt disappear, remove it and remodel to ensure it doesnt come back.

[27-Sep-2011 09:06:25] <humanix> just like that? proc

[27-Sep-2011 09:06:33] <Hackman238> humanix: I think so.

[27-Sep-2011 09:07:37] <Hackman238> humanix: Acctually put this instead: ^\/dev|^\/proc

[27-Sep-2011 09:08:11] <Hackman238> humanix: That'll catch all /dev block devices and /proc. It wont act on mounted filesystems

[27-Sep-2011 09:08:22] <Hackman238> *act on other mounted filesystems

[27-Sep-2011 09:08:41] <humanix> like a regex

[27-Sep-2011 09:08:52] <Hackman238> humanix: Yes

[27-Sep-2011 09:09:19] <Hackman238> enkrypt: This StackSub_St1-1, what is it?

[27-Sep-2011 09:10:51] <Hackman238> enkrypt: I'm thinking most likely it doesnt support most counters. you can add it to zInterfaceMapIgnoreNames as ^(StackSub\d+)$

[27-Sep-2011 09:11:11] <Hackman238> enkrypt: Remodel and see if it goes away. If it doesnt, remove it and remodel to ensure it doesnt come back

[27-Sep-2011 09:16:55] <jmp242> Hackman238: do you think it would be too much if we had a sig or something that indicated these responses were under the auspices of the ZCA? A little promotion?

[27-Sep-2011 09:17:09] <Hackman238> jmp242: Not at all.

[27-Sep-2011 09:17:31] <Hackman238> jmp242: "You benovolent masters, the ZCA" XD

[27-Sep-2011 09:17:46] <Hackman238> jmp242: Any idea what would be good?

[27-Sep-2011 09:17:55] <Hackman238> s/you/your/

[27-Sep-2011 09:18:37] <jmp242> well, I was going to say something like

[27-Sep-2011 09:18:45] <jmp242> answered by the ZCA. I sent out a mailing on it

[27-Sep-2011 09:18:59] <jmp242> I haven't seen us ever get the badges like zenmaster or whatever

[27-Sep-2011 09:19:12] <jmp242> so I'm going to see if jive can let us do a sig, or if I have to keep pasting it into the forums

[27-Sep-2011 09:19:22] <jmp242> btw - e-mail forum replies are still bounced to me

[27-Sep-2011 09:19:24] <jmp242>

[27-Sep-2011 09:19:31] <Hackman238> jmp242: Alrighty.

[27-Sep-2011 09:19:35] <Hackman238> jmp242: Good idea!

[27-Sep-2011 09:19:44] <Hackman238> jmp242: Yes...mine too. I told Nick.

[27-Sep-2011 09:20:21] <Hackman238> jmp242: He's screwed for time...hes been too busy to even get back to me in Skype lately

[27-Sep-2011 09:22:01] <jmp242> wow

[27-Sep-2011 09:22:13] <jmp242> I hope that's good ... ~?

[27-Sep-2011 09:22:40] <Hackman238> jmp242: Well yes and no.

[27-Sep-2011 09:23:05] <Hackman238> jmp242: He's trying to do a lot for us, but at the same time he needs an assistant.

[27-Sep-2011 09:23:54] <ion-fan> hackman238 can you take a look at my forums post please? (message/61547#61547)

[27-Sep-2011 09:24:13] <Hackman238> jmp242: Zenoss is about as organized management wise as a group of parkinson's patients collating paperwork.

[27-Sep-2011 09:24:49] <Hackman238> jmp242: They have good intentions, they just don't have enough staff right now so it's a bit crazy.

[27-Sep-2011 09:24:52] <jmp242> ion-fan: that's on the ZCA list

[27-Sep-2011 09:24:57] <rocket> Hackman238: really?

[27-Sep-2011 09:25:19] <jmp242> Hackman238: yea I get the feeling like they flail around a bit wrt Core

[27-Sep-2011 09:25:23] <jmp242> but that has been explained

[27-Sep-2011 09:25:28] <Hackman238> rocket: That's the excuse everyone overthere management bottom to top gives me.

[27-Sep-2011 09:25:30] <jmp242> I can't speak to Enterprise

[27-Sep-2011 09:25:33] <Hackman238> *over there

[27-Sep-2011 09:26:16] <ion-fan> sorry but what is the zca list?

[27-Sep-2011 09:26:18] <Hackman238> jmp242: They're working on it.

[27-Sep-2011 09:27:23] <enkrypt> Hackman238: thanks, I simply added "StackSub" to the end of my Interface Ignore list, and it went away. was just a bogus interface which has half of the regular OID's missing

[27-Sep-2011 09:28:18] <Hackman238> ion-fan: Yikes- thats a crazy problem. Transforms shouldn't have anything to do with that functionality.

[27-Sep-2011 09:28:20] <rocket> well we dont have enough staff, but really the generalizations like that do not help foster trust

[27-Sep-2011 09:28:39] <jmp242> ion-fan: The ZCA has an informal list of posts on the forums that we want to help out with

[27-Sep-2011 09:28:44] <Hackman238> rocket: That's what I said.

[27-Sep-2011 09:29:15] <jmp242> I generally maintain it, and Hackman238 and other ZCA members help out with the tricky ones

[27-Sep-2011 09:30:02] <jmp242> it's part of our pledge here: community/zca/blog/2011/08/26/zca-8-26-2011

[27-Sep-2011 09:30:34] <ion-fan> i want to know whats behind that abbreviation Zenoss Community A... ???

[27-Sep-2011 09:30:47] <jmp242> Alliance

[27-Sep-2011 09:31:39] <ion-fan> aah

[27-Sep-2011 09:32:13] <JohnnyNOC> is there a stable release of 3.2?

[27-Sep-2011 09:32:21] <jmp242> https://zcaportal.org/wiki/bin/view/Main/WebHome

[27-Sep-2011 09:32:32] <jmp242> see our fledgling wiki

[27-Sep-2011 09:32:53] <jmp242> JohnnyNOC: there is

[27-Sep-2011 09:32:59] <jmp242> but you should install the listed patches

[27-Sep-2011 09:33:14] <jmp242> I believe 3.2.1 will be out soonish with those included

[27-Sep-2011 09:33:23] <JohnnyNOC> hmm..  i'll probably just wait

[27-Sep-2011 09:33:28] <JohnnyNOC> not worth having things blow up on me

[27-Sep-2011 09:33:29] <JohnnyNOC>

[27-Sep-2011 09:33:39] <ion-fan> so you mean that the global search is going to be implemented in the next version ? through the work of zca??

[27-Sep-2011 09:33:48] <dhopp> JohnnyNOC: 3.2.1 is in beta right now which has all of the critical patches plus some other ones.  I haven't seen or heard any major issues with 3.2.1 beta so it should get officially released relatively soon

[27-Sep-2011 09:36:31] <JohnnyNOC> what version are you running dhopp

[27-Sep-2011 09:37:45] <dhopp> in the new production setup that I'm building I've installed 3.2.0 with the patches

[27-Sep-2011 09:39:11] jellis is now known as poised

[27-Sep-2011 09:39:12] poised is now known as jellis

[27-Sep-2011 09:41:31] <humanix> thans Hackman238

[27-Sep-2011 09:41:31] <humanix>

[27-Sep-2011 09:42:05] <JohnnyNOC> Hackman238 i almost signed up on the ZCA wiki as Tito Ortiz

[27-Sep-2011 09:42:09] <JohnnyNOC> just to troll you

[27-Sep-2011 09:42:17] <JohnnyNOC> first post title - "Shane Scott is Dreamy!"

[27-Sep-2011 09:42:21] <JohnnyNOC> but i got lazy

[27-Sep-2011 09:42:22] <JohnnyNOC> :/

[27-Sep-2011 09:42:38] <Hackman238> JohnnyNOC: LOL!

[27-Sep-2011 09:42:44] <Hackman238> humanix: NP

[27-Sep-2011 09:43:02] <Hackman238> ion-fan: It may be implemented in v4.

[27-Sep-2011 09:43:10] <Hackman238> ion-fan: I'll investigate the error today

[27-Sep-2011 09:43:15] <Hackman238> enkrypt: NP

[27-Sep-2011 09:46:01] <ion-fan> hackman238: ok thank you

[27-Sep-2011 09:50:13] <dhopp> Hackman238:  in your opinion where is about the threshold of devices before breaking a physical collector into logical collectors makes sense?

[27-Sep-2011 09:52:41] <Hackman238> dhopp: Yes. Consider the average time it takes to poll one of the devices in its entirety and add the snmp retries x snmp timeout value x the total count of devices. It needs to fit under the polling cycle period

[27-Sep-2011 09:53:37] <Hackman238> dhopp: so- 'Average Polling time' + ('snmp timout' x 'snmp retries') x 'Total devices'

[27-Sep-2011 09:54:05] <Hackman238> dhopp: This leaves room when SHTF and devices stop replying quickly or at all.

[27-Sep-2011 09:55:32] <Hackman238> dhopp: It has less to do with the number of devices and more with the number of datapoints. By taking an average of the polling time you can decern about how long each device is really contributing to the cycle time.

[27-Sep-2011 09:55:37] <dhopp> Hackman238:  I'm trying to decide if I do logical collectors now or later.  My physical collectors are 16 core so I want to be able to utilize them and was thinking of doing 2 logical, but don't want to overcomplicate the setup either…each physical collector will probably have ~600 or so devices which I think could be handled without logical (5 minute polling for snmp)

[27-Sep-2011 09:56:08] <Hackman238> dhopp: Nice boxes, but unless you are using logical collectors most of those cores go unused.

[27-Sep-2011 09:59:05] <dhopp> Hackman238:  that's what I figured

[27-Sep-2011 09:59:21] <dhopp> although we will have a good number of command datasources

[27-Sep-2011 10:00:02] <dhopp> Hackman238:  we will have a lot of check_http and check_apache along with some other commands that will have to run

[27-Sep-2011 10:00:14] <Hackman238> dhopp: Careful with commadn datasources- those can be scaled using the parallel option, but they can pile up like 1604 in San Antonio when it rains.

[27-Sep-2011 10:00:24] <JohnnyNOC> hah

[27-Sep-2011 10:00:30] <JohnnyNOC> nice analogy

[27-Sep-2011 10:00:36] <Hackman238> JohnnyNOC: Thanks!

[27-Sep-2011 10:00:37] <JohnnyNOC> nyeates in da houzzz

[27-Sep-2011 10:00:51] <dhopp> Hackman238: don't even get me started on rain and traffic…I don't think there is anything more aggrovating in life then driving in Chicago when it rains or snows

[27-Sep-2011 10:01:00] <Hackman238> dhopp: I would highly suggest using the http test pack instead of check_http

[27-Sep-2011 10:01:08] <JohnnyNOC> dhopp there are worse, believe it or not

[27-Sep-2011 10:01:18] <dhopp> It's fricken CHICAGO..it rains or snows like 6 months out of the year but nobody around here knows how to drive in it

[27-Sep-2011 10:01:22] <Hackman238> nyeates: Hey, how goes it?

[27-Sep-2011 10:01:29] <Hackman238> dhopp: Madness!

[27-Sep-2011 10:01:32] <JohnnyNOC> they're from the burbs

[27-Sep-2011 10:01:34] <JohnnyNOC> thats what i say

[27-Sep-2011 10:01:48] <nyeates> heya guys

[27-Sep-2011 10:01:55] <JohnnyNOC> hey youuuuuu guyyyyyyys

[27-Sep-2011 10:02:05] <JohnnyNOC> what movie

[27-Sep-2011 10:02:14] <Hackman238> nyeates: Hows it going?

[27-Sep-2011 10:02:15] <nyeates> working non-stop on scripting up something to bring zenpacks over to individual repos on git

[27-Sep-2011 10:02:17] <nyeates> GOONIES

[27-Sep-2011 10:02:24] <JohnnyNOC> hahaha nice nyeates

[27-Sep-2011 10:02:42] <Hackman238> dhopp: Command datasources help make zenoss flexible, but you definately want to minimize their use much as possible

[27-Sep-2011 10:02:49] <Hackman238> nyeates: LOL

[27-Sep-2011 10:02:56] <Hackman238> nyeates: Very cool

[27-Sep-2011 10:02:58] <dhopp> JohnnyNOC: I grew up in Northern Michigan and when I moved down here and the first snow 'storm' schools were closed and everybody called into work..I drove my front wheel drive saturn and was like 'why is everybody worried'...

[27-Sep-2011 10:03:14] <nyeates> Goonies is cult classic of the 80s

[27-Sep-2011 10:03:33] <dhopp> Hackman238:  I don't think we will have thousands by any stretch…but there will be a good amount running

[27-Sep-2011 10:03:34] <JohnnyNOC> dhopp did you see pics of the cars stranded on LSD last year?

[27-Sep-2011 10:03:41] <JohnnyNOC> i think its stuff like that that freaks people out

[27-Sep-2011 10:03:41] <JohnnyNOC>

[27-Sep-2011 10:03:43] <dhopp> JohnnyNOC: that was actually a storm

[27-Sep-2011 10:03:53] <JohnnyNOC> well yea

[27-Sep-2011 10:04:00] <dhopp> 1-2" of snow fall is not a storm

[27-Sep-2011 10:04:18] <Hackman238> dhopp: Nope. 6"+ is a storm

[27-Sep-2011 10:04:20] <Hackman238>

[27-Sep-2011 10:04:21] <nyeates> cars on LSD... did i read that right?

[27-Sep-2011 10:04:28] <Hackman238> nyeates: bridge

[27-Sep-2011 10:04:31] <dhopp> nyeates:  lol…he meant lake shore drive

[27-Sep-2011 10:04:46] <nyeates> hah

[27-Sep-2011 10:04:51] <JohnnyNOC> http://blog.pictopia.com/wp-content/uploads/2011/02/59171631.jpg

[27-Sep-2011 10:04:59] <nyeates> WHOA

[27-Sep-2011 10:05:05] <JohnnyNOC> that's a "highway"

[27-Sep-2011 10:05:06] <JohnnyNOC> ?

[27-Sep-2011 10:05:06] <Hackman238> http://en.wikipedia.org/wiki/Lake_Shore_Drive

[27-Sep-2011 10:05:10] <nyeates> looks like a post apocalypitic picture

[27-Sep-2011 10:05:16] <Hackman238> nyeates: LOL

[27-Sep-2011 10:05:21] <JohnnyNOC> people got sick of waiting, or they ran out of gas, or their cell phones ran out

[27-Sep-2011 10:05:25] <JohnnyNOC> so thye left their cars on the road

[27-Sep-2011 10:05:37] <dhopp> nyeates:  yeah it was crazy…there were 3 lane roads that were only 2 lanes for like a week because they couldn't clear all of the snow so they just piled it up in the right lane

[27-Sep-2011 10:06:03] <nyeates> wh

[27-Sep-2011 10:06:53] <Hackman238> dhopp: Any way....try to avoid becoming reliant on command datasources now while your zenoss is smaller and easier to restructure

[27-Sep-2011 10:07:36] <dhopp> Hackman238:  where is the http test pack?  or is that an enterprise only ting?

[27-Sep-2011 10:07:37] <dhopp> thing

[27-Sep-2011 10:08:02] <Hackman238> dhopp: docs/DOC-3431

[27-Sep-2011 10:09:18] <dhopp> that's what I'm using..but I thought that did a command datasource

[27-Sep-2011 10:10:32] <nyeates> any of you hardcore zenpack dev's use an IDE? (Integrated Development Environment)

[27-Sep-2011 10:10:54] <nyeates> or advanced text editors?

[27-Sep-2011 10:11:13] <JohnnyNOC> i've been messing with eclipse, but all my dev for zenoss seems to happen remotely at the command line

[27-Sep-2011 10:11:19] <JohnnyNOC> so i've been thinkin about pimping aobut my .vimrc

[27-Sep-2011 10:11:41] <JohnnyNOC> but i'm not much of a developer

[27-Sep-2011 10:11:52] <dhopp> nyeates:  I'm not hard core (just starting)..but I played around with using an IDE a little bit and couldn't deal with the pain in the ass of setting up the environment to take advantage of the IDE features…so have fallen back to trusty vi

[27-Sep-2011 10:12:16] <nyeates> Yeah, someone had mentioned Eclipse

[27-Sep-2011 10:12:22] <JohnnyNOC> i just need syntax highliting and tab completion

[27-Sep-2011 10:12:24] <JohnnyNOC> which should be doable with vim

[27-Sep-2011 10:12:48] <nyeates> Many here use textmate on mac....or like Johnny said, pimped out VI or VIM

[27-Sep-2011 10:13:21] <JohnnyNOC> nyeates why do you ask?

[27-Sep-2011 10:14:07] <nyeates> Employee from IBMs Linux Research Center, who made the latest zenpack, was asking questions about Eclipse IDE....and i was curious what others have used

[27-Sep-2011 10:14:18] <JohnnyNOC> ah

[27-Sep-2011 10:14:21] <JohnnyNOC> nyeates they made that zenpack for us

[27-Sep-2011 10:14:23] <JohnnyNOC> jsyk

[27-Sep-2011 10:14:36] <nyeates> jsyk?

[27-Sep-2011 10:14:40] <JohnnyNOC> i been trying to find people who use the IBM IMM's to test it out and stuff..  no luck however

[27-Sep-2011 10:14:44] <JohnnyNOC> jsyk = just so you know

[27-Sep-2011 10:15:19] <nyeates> JohnnyNOC: oh wow really? I did not know that. Yeah he said they made it for a client. Wow. Congrats that it worked out

[27-Sep-2011 10:15:36] <JohnnyNOC> well, it's far from perfect

[27-Sep-2011 10:15:58] <JohnnyNOC> but we can't run their ibm director agent on our linux machines (debian based) so they put that together for us so we have some monitoring

[27-Sep-2011 10:16:00] <JohnnyNOC> snmp traps

[27-Sep-2011 10:16:03] <nyeates> I am surprised IBM agreed to dev that.... did they have any prior workings or knowledge with zenoss?

[27-Sep-2011 10:16:07] <JohnnyNOC> no

[27-Sep-2011 10:16:10] <JohnnyNOC> they want us to buy servers

[27-Sep-2011 10:16:18] <JohnnyNOC> and said thats one of the things they would do to try and win our business

[27-Sep-2011 10:16:45] <nyeates> Huh...very interesting.... they must have tight competition....or you guys are waving big bucks :-)

[27-Sep-2011 10:16:57] <JohnnyNOC> haha

[27-Sep-2011 10:16:59] <JohnnyNOC> those things aren't cheap

[27-Sep-2011 10:17:31] enkrypt_ is now known as enkrypt

[27-Sep-2011 10:18:16] <JohnnyNOC> but yea, if you know anyone using ibms with imms ask them to try it out

[27-Sep-2011 10:18:20] <JohnnyNOC> i want to get them as much feedback ya know

[27-Sep-2011 10:19:37] <jmp242> I use IBMs - some anywau

[27-Sep-2011 10:19:48] <jmp242> but I don't know that I want to try a beta Zenback

[27-Sep-2011 10:19:51] <jmp242> err

[27-Sep-2011 10:19:52] <jmp242> Zenpack

[27-Sep-2011 10:20:08] <jmp242> OTOH, I wasn't going to run systems director eitehr

[27-Sep-2011 10:20:12] <jmp242> and I can't type today

[27-Sep-2011 10:20:49] <jmp242> Unfortunately we would have no pull with IBM lol - we only buy 2 or 3 servers at a time

[27-Sep-2011 10:22:02] <nyeates> jmp242: try it on a dev or test zenoss box - if you have one

[27-Sep-2011 10:22:23] <jmp242> If I had one lol

[27-Sep-2011 10:23:13] <nyeates> ;-)

[27-Sep-2011 10:35:02] <dhopp> jmp242:  dev/test box is good…prevents you from having to rely on backups so much ;-)

[27-Sep-2011 10:35:31] <lavalamp> morning

[27-Sep-2011 10:35:31] <jmp242> yea

[27-Sep-2011 10:35:43] <dhopp> nyeates:  how often should zeopack be ran?  is every 4 hours too much?

[27-Sep-2011 10:35:43] <jmp242> I just got the OK to ask our CIT to see what they charge for a hosted VM

[27-Sep-2011 10:36:55] lavalamp is now known as bseklecki

[27-Sep-2011 10:52:17] <enkrypt> rmatte: I'm using your Windows CPU+RAM simple zenpack to monitor 1154 windows machines without a hitch only too bad we can't do the same for windows 2000

[27-Sep-2011 10:55:32] <jmp242> enkrypt: Windows 2000? EOL it lol

[27-Sep-2011 10:56:42] <enkrypt> jmp242: yeah it's no biggie, from the 1154 windows servers I see only 23 of them are running 2000

[27-Sep-2011 10:57:19] <enkrypt> but I'm glad rmatte's zenpack exists, because we don't use Informant SNMP, and it's not possible for us to use WMI on each of these servers

[27-Sep-2011 10:57:32] <enkrypt> so at least this gives us basic CPU & ram graphs

[27-Sep-2011 10:58:02] <Jane_Curry> Hi guys (and gals if there are any out there    )

[27-Sep-2011 10:58:35] <jbk> is tehre a decent WMI browser to be able to find out what things can be monitored on a windows box?

[27-Sep-2011 10:59:09] <jbk> for example, it'd be _really_ nice to look at AD replication stats to spot and DCs that are having replication issues

[27-Sep-2011 10:59:38] <Jane_Curry> jbk: I just have some saved wmic commands that I use to test WMI but a heterogeneous WMI browser would be useful (ie one that runs on Linux as well as Doze)

[27-Sep-2011 10:59:53] <jbk> well even on windows

[27-Sep-2011 11:00:12] <jbk> just so i can figure out what can be setup in zenoss

[27-Sep-2011 11:00:38] <enkrypt> jbk: I would be interested in that as well

[27-Sep-2011 11:01:14] <Jane_Curry> Can anyone comment on the scale of their Zenoss Core installations???

[27-Sep-2011 11:01:55] <enkrypt> Jane_Curry: running enterprise, sorry

[27-Sep-2011 11:01:56] <dhopp> this doesn't work for wmi browsing? http://www.microsoft.com/download/en/details.aspx?id=24045 (of course doesn't run on linux)

[27-Sep-2011 11:01:57] <Jane_Curry> I have a potential client who is talking about scaling to 5000 devices in the next 18 months and upto 10,000 in 2 years - is this too big for Core???

[27-Sep-2011 11:02:18] <dpetzel> Jane_Curry: We are monitoring just under 700 devices (mostly servers) with 1 master/hub two remote collectors

[27-Sep-2011 11:02:42] <Jane_Curry> Devices are largely fairly simple Wifi access devices - simple availability monitoring and a few SNMP templates

[27-Sep-2011 11:02:56] <enkrypt> Jane_Curry: you will need a lot of hardware for that, and certainly several hubs and distributed collectors. afaik distributed collectors are not included in Core?

[27-Sep-2011 11:03:11] <Jane_Curry> Thanks dpetzel - good starting point - no problems??

[27-Sep-2011 11:03:42] <dpetzel> Jane_Curry:  No problems so far, we really dont need the extra collector for load, it was a firewall/scurity decision to monitor a small subset of protected hosts

[27-Sep-2011 11:04:13] <dhopp> Jane_Curry:  there is a forum post were some people are monitoring 7000 devices but they have spent a lot of time tweaking their setups…I'm working on installing Core to monitor ~1200 devices (about 800 servers, 400 network devices)

[27-Sep-2011 11:04:15] <dpetzel> Jane_Curry: our main collector manages about 600 of those devices, and we've not seen any real issues

[27-Sep-2011 11:04:37] <dhopp> dpetzel:  what's the specs on that collector?

[27-Sep-2011 11:04:47] <Jane_Curry> I believe Distributed Collectors are possible in Core - I lab-tested it at one time and one of the cognoscenti has written a distributed collector Core ZenPack

[27-Sep-2011 11:05:16] <nyeates> Any issues with 3.2.1 beta peoples?

[27-Sep-2011 11:05:25] <nyeates> anyone had a chance to test at it?

[27-Sep-2011 11:05:28] <dhopp> Jane_Curry:  distributed collectors are definitely possible (egor's pack is pretty good) and you can do logical collectors (more than one collector / physical collector) but that requires some manual setup

[27-Sep-2011 11:05:40] <dhopp> nyeates:  only basic testing so far and don't see a problem

[27-Sep-2011 11:06:01] <dpetzel> dhopp: Dell M610LG with I believe 2 quad cores, and 16GB of RAM (trying to locate the exact info)

[27-Sep-2011 11:06:06] <enkrypt> oh ok, in that case you can work it out I guess. at work I'm doing a project to monitor 2300 assets, growing to 3000 quickly. 1 hub, 4 collectors (well 2 active collectors in fact, because I have 2 active-passive high-availability pairs, using DRBD and heartbeat)

[27-Sep-2011 11:06:35] <nyeates> dhopp: thanks for the feedback

[27-Sep-2011 11:06:47] <nyeates> I think no news is good news here

[27-Sep-2011 11:08:24] <nyeates> We will leave beta 3.2.1 open for testing until at least end of this week... check it out if you get a chance: blogs/zenossblog/2011/09/23/release-of-beta-2-for-321--added-stack-and-vmware-installs

[27-Sep-2011 11:09:32] <dpetzel> dhopp:  I lied, its our larger sku on this main collector so its dual-quad (Xeon X5550 2.67ghz) with 96GB of ram not 16

[27-Sep-2011 11:10:00] <Hackman238> nyeates: I haven't had a chance. Been evaluating ZEC/GOM

[27-Sep-2011 11:10:47] <dhopp> Hackman238:  ZEC/GOM?

[27-Sep-2011 11:14:19] <Hackman238> dhopp: Event Consolidator, its an enterprise thing

[27-Sep-2011 11:25:36] <nyeates> Zenoss Event Consolidator / Global Opertions Manager

[27-Sep-2011 11:25:50] <nyeates> ZEC was the old name... GOM is the new name

[27-Sep-2011 11:26:22] <nyeates> allows you to bring events from multiple zenoss instances, into one

[27-Sep-2011 11:27:14] <dhopp> nyeates:  I'm looking at a 3.2.0 install right now..and can't seem to be able to override a template when looking at a device or device class under the infrastructure view

[27-Sep-2011 11:27:39] <Hackman238> ion-fan: Can you run a zenchkrels -x1 -r for me?

[27-Sep-2011 11:28:09] <Hackman238> ion-fan: Also, did you have the zenoss core device search already installed?

[27-Sep-2011 11:28:24] <ion-fan> yes but it didnt work

[27-Sep-2011 11:28:55] <ion-fan> zenchkrels -x1 -r  as zenoss user?

[27-Sep-2011 11:30:17] <ion-fan> whats behind that command?

[27-Sep-2011 11:31:18] <Hackman238> ion-fan: yep as zenoss user. It'll go through and check relationships

[27-Sep-2011 11:31:59] <Hackman238> ion-fan: What I mean is if you had the zenoss core device search installed already you'll need to remove it then reinstall my pack

[27-Sep-2011 11:32:05] <Hackman238> ion-fan: the two packs conflict

[27-Sep-2011 11:32:21] <ion-fan> do you really need all the information that the command spites out?

[27-Sep-2011 11:32:37] <Hackman238> ion-fan: No, only if it stops and show traceback

[27-Sep-2011 11:32:44] <ion-fan> ah ok

[27-Sep-2011 11:32:51] <ion-fan> as file or post it here

[27-Sep-2011 11:33:12] <ion-fan> (about 10 rows)

[27-Sep-2011 11:33:21] <Hackman238> ion-fan: Either

[27-Sep-2011 11:33:31] <Hackman238> ion-fan: post is fine, easier

[27-Sep-2011 11:33:39] <ion-fan> ERROR:zen.Relations:remote relation /zport/dmd/ZenUsers/admin/messages/msg/messageQueue doesn't point back to /zport/dmd/ZenUsers/admin/messages/msg/messageQueue

[27-Sep-2011 11:33:39] <ion-fan> WARNING:zen.Relations:reconnecting relation /zport/dmd/ZenUsers/admin/messages/msg/messageQueue to relation /zport/dmd/ZenUsers/admin/messages/msg/messageQueue

[27-Sep-2011 11:33:39] <ion-fan> Traceback (most recent call last):

[27-Sep-2011 11:33:39] <ion-fan>   File "/usr/local/zenoss/zenoss/Products/ZenUtils/CheckRelations.py", line 72, in <module>

[27-Sep-2011 11:33:39] <ion-fan>     tmbk.rebuild()

[27-Sep-2011 11:33:40] <ion-fan>   File "/usr/local/zenoss/zenoss/Products/ZenUtils/CheckRelations.py", line 40, in rebuild

[27-Sep-2011 11:33:40] <ion-fan>     object.checkRelations(repair=repair)

[27-Sep-2011 11:33:41] <ion-fan>   File "/var/zenoss/zenoss/Products/ZenRelations/RelationshipManager.py", line 262, in checkRelations

[27-Sep-2011 11:33:41] <ion-fan>     rel.checkRelation(repair)

[27-Sep-2011 11:33:42] <ion-fan>   File "/var/zenoss/zenoss/Products/ZenRelations/ToOneRelationship.py", line 223, in checkRelation

[27-Sep-2011 11:33:42] <ion-fan>     rrel._add(parobj)

[27-Sep-2011 11:33:43] <ion-fan>   File "/var/zenoss/zenoss/Products/ZenRelations/ToOneRelationship.py", line 80, in _add

[27-Sep-2011 11:33:43] <ion-fan>     self._remoteRemove()

[27-Sep-2011 11:33:44] <ion-fan>   File "/var/zenoss/zenoss/Products/ZenRelations/ToOneRelationship.py", line 102, in _remoteRemove

[27-Sep-2011 11:34:08] <Hackman238> ion-fan: Okay theres some stuff afoot.

[27-Sep-2011 11:34:16] <Hackman238> ion-fan: As zenoss user, zendmd

[27-Sep-2011 11:34:39] <ion-fan> jep done

[27-Sep-2011 11:35:00] <Hackman238> ion-fan: paste for user in dmd.ZenUsers.getUsers(): user.messages.removeRelation()

[27-Sep-2011 11:35:02] <ion-fan> alot of warnings appear before i get the command prompt

[27-Sep-2011 11:35:06] <Hackman238> ion-fan: {double enter}

[27-Sep-2011 11:35:10] <Hackman238> ion-fan: commit()

[27-Sep-2011 11:35:14] <Hackman238> ion-fan: What are the errors?

[27-Sep-2011 11:36:24] <ion-fan> done no errors

[27-Sep-2011 11:36:35] <ion-fan> nothing happened

[27-Sep-2011 11:36:38] <Hackman238> ion-fan: okay, commit()

[27-Sep-2011 11:36:44] <ion-fan> yes already done

[27-Sep-2011 11:36:59] <Hackman238> ion-fan: okay good. It shouldnt output anything unless something really is wrong

[27-Sep-2011 11:37:11] <Hackman238> ion-fan: CTRL + D to escape

[27-Sep-2011 11:37:16] <Hackman238> ion-fan: vi %ZENHOME/ZenPacks/ZenPacks.ShaneScott.DeviceSearch-1.0.0-py2.6.egg/ZenPacks/ShaneScott/DeviceSearch/__init__.py

[27-Sep-2011 11:37:44] <Hackman238> ion-fan: chop out that extra code you added and save it. We'll readd once the errors are gone

[27-Sep-2011 11:39:23] <ion-fan> done

[27-Sep-2011 11:40:18] <ion-fan> it was only device.os.interfaces.interface.description at the end of row 30...(in the brackets)

[27-Sep-2011 11:43:20] <dhopp> anybody know if DNSMonitor template can look to make sure a resolution resolves to a CNAME?  and if so, can use a regex to match the CNAME?

[27-Sep-2011 11:47:40] <Hackman238> ion-fan: Okay. ideally we want to limit the number of possabilities causing the problem.

[27-Sep-2011 11:48:21] <Hackman238> ion-fan: After switching that _init_.py file back, issue a zopectl restart

[27-Sep-2011 11:48:38] <Hackman238> ion-fan: That'll restart zope to reload the pack without interrupting monitoring

[27-Sep-2011 11:48:47] <ion-fan> you mean i have to put that original file over the current file

[27-Sep-2011 11:48:56] <Hackman238> ion-fan: Yep

[27-Sep-2011 11:49:32] <Hackman238> ion-fan: I just want to confirm that ZenPacks.zenoss.DeviceSearch is not installed, is that right?

[27-Sep-2011 11:50:29] <ion-fan> no it is installed

[27-Sep-2011 11:50:33] <Hackman238> ion-fan: Gotcha

[27-Sep-2011 11:50:59] <ion-fan> otherwise there cant be this __init... file ?

[27-Sep-2011 11:51:56] <Hackman238> ion-fan: ZenPacks.zenoss.DeviceSearch is a pack that does some of the same things ZenPacks.ShaneScott.DeviceSearch does, so they will conflict if installed together.

[27-Sep-2011 11:52:33] <Hackman238> ion-fan: That _init_.py file must be present, I just want to switch it back to its unedited state to ensure that isnt part of the problem.

[27-Sep-2011 11:54:58] <ion-fan> ah i didnt look properly... but anyhow it was never installed ...

[27-Sep-2011 11:55:20] <ion-fan> zenoss@mycroft:/usr/local/zenoss/zenoss/ZenPacks$ ls

[27-Sep-2011 11:55:20] <ion-fan> easy-install.pth  ZenPacks.ShaneScott.DeviceSearch-1.0.0-py2.6.egg

[27-Sep-2011 11:55:37] <Hackman238> ion-fan: Oh that wont work

[27-Sep-2011 11:55:45] <ion-fan> thats the content of the Zenpack directory

[27-Sep-2011 11:55:54] <Hackman238> ion-fan: you'll need to su zenoss; zenpack --install ZenPacks.ShaneScott.DeviceSearch-1.0.0-py2.6.egg

[27-Sep-2011 11:56:04] <Hackman238> Oh ok

[27-Sep-2011 11:56:07] <ion-fan> yes i have done that

[27-Sep-2011 11:56:25] <ion-fan> at the beginning of my log post

[27-Sep-2011 11:56:53] <Hackman238> ion-fan: Oh ok. Lets reinstall this puppy.

[27-Sep-2011 11:57:18] <ion-fan> seems to be a good solution^^

[27-Sep-2011 11:57:25] <Hackman238> ion-fan: su zenoss; zenpack --remove ZenPacks.ShaneScott.DeviceSearch; zopectl restart; zenpack --install ZenPacks.ShaneScott.DeviceSearch-1.0.0-py2.6.egg

[27-Sep-2011 11:58:54] <ion-fan> the remove command also triggers these errors i get when i installed it

[27-Sep-2011 11:59:05] <Hackman238> ion-fan: okay, lets look before we continue

[27-Sep-2011 11:59:20] <Hackman238> ion-fan: ls $ZENHOME/ZenPacks

[27-Sep-2011 11:59:26] <Hackman238> ion-fan: Is the device search pack in there yet?

[27-Sep-2011 12:00:01] [disconnected at Tue Sep 27 12:00:01 2011]

[27-Sep-2011 12:00:02] [connected at Tue Sep 27 12:00:02 2011]

[27-Sep-2011 12:00:14] [zenoss-logger (logger bot) has joined #zenoss]

[27-Sep-2011 12:00:21] <ion-fan> nope

[27-Sep-2011 12:00:42] <ion-fan> zenoss@mycroft:/usr/local/zenoss/zenoss/ZenPacks$ ls $ZENHOME/ZenPacks

[27-Sep-2011 12:00:42] <ion-fan> ZenPacks.ShaneScott.DeviceSearch-1.0.0-py2.6.egg

[27-Sep-2011 12:00:42] <ion-fan> zenoss@mycroft:/usr/local/zenoss/zenoss/ZenPacks$

[27-Sep-2011 12:01:08] <Hackman238> ion-fan: Okay, rm that ZenPacks.ShaneScott.DeviceSearch-1.0.0-py2.6.egg directory

[27-Sep-2011 12:01:24] <Hackman238> ion-fan: I tihnk I found a bug. Give me 5

[27-Sep-2011 12:02:10] <ion-fan> ok

[27-Sep-2011 12:02:17] <ion-fan> delete done

[27-Sep-2011 12:02:47] <Hackman238> ion-fan: Alrighty. Whats your email address?

[27-Sep-2011 12:02:53] <Hackman238> ion-fan: pm me if you want

[27-Sep-2011 12:04:15] <ion-fan> michael@work.de

[27-Sep-2011 12:04:27] <ion-fan> its my company mail address^^

[27-Sep-2011 12:06:11] <dhopp> Hackman238:  you have any ideas about this http://pastebin.com/wJpLmDRT

[27-Sep-2011 12:06:29] <dhopp> it seems to have just started

[27-Sep-2011 12:07:01] <Hackman238> ion-fan: I jsut emailed up a modified copy. I think part of the reason is by default the pack is indexing properties that may not exist on your system

[27-Sep-2011 12:07:58] <Hackman238> dhopp: try bouncing the zenhub that feeds that zenperfsnmp, then test using zenperfsnmp run now -v10 -d {broke device}

[27-Sep-2011 12:08:15] <dhopp> I'm not sure what device is broken

[27-Sep-2011 12:08:15] <Hackman238> dhopp: If still fail, su zenoss; zenchkrels -x1 -r

[27-Sep-2011 12:08:29] <Hackman238> dhopp: chi-zen-dev-02_net01 isnt a device?

[27-Sep-2011 12:08:44] <dhopp> it's a logical collector "net01"

[27-Sep-2011 12:08:52] <Hackman238> dhopp: Ah gotcha

[27-Sep-2011 12:08:59] <dhopp> I think this happened when I did a "Push Changes" from a device

[27-Sep-2011 12:09:14] <Hackman238> dhopp: I'd bounce the hub that feeds that collector and all the colelctors daemons

[27-Sep-2011 12:09:20] <Hackman238> dhopp: then retry the push

[27-Sep-2011 12:09:34] <Hackman238> dhopp: it could be a relationship issue if that doesnt fix it

[27-Sep-2011 12:10:02] <Hackman238> ion-fan: Once that new pack is downloaded, su zenoss; zenpack --install ZenPacks.ShaneScott.DeviceSearch-1.0.0-py2.6.egg

[27-Sep-2011 12:10:02] <rocket> it is a relationship issue

[27-Sep-2011 12:10:06] <Hackman238> ion-fan: Any errors?

[27-Sep-2011 12:10:19] <rocket> the link between the hub and collector definition is borked

[27-Sep-2011 12:10:31] <dhopp> rocket: how do I fix it?

[27-Sep-2011 12:10:39] <Hackman238> dhopp: if it is a relationship issue we can try to fix it with zenchkrels -x1 -r

[27-Sep-2011 12:10:50] <rocket> try zenchkrels -x1 -r first

[27-Sep-2011 12:11:19] <Hackman238> rocket: Sorry, I didnt realize you chimed in

[27-Sep-2011 12:11:31] <Hackman238> rocket: I just kept typing LOL

[27-Sep-2011 12:11:32] <rocket> no problem

[27-Sep-2011 12:14:31] <dhopp> I ran zenchkrels but I don't see anything that indicates whether if found a problem or not (just a crap load of INFO messages)

[27-Sep-2011 12:16:04] <rocket> can you remove the collector from the ui and then add it back?

[27-Sep-2011 12:17:03] <dhopp> I just did a push changes on the device that I think was the problem and it didn't bomb out this time….

[27-Sep-2011 12:17:34] <dhopp> right before that trace back was: 2011-09-22 10:49:23,703 INFO zen.zenperfsnmp: Checking for outdated configs..

[27-Sep-2011 12:18:04] <rocket> so its resolved?

[27-Sep-2011 12:18:24] <ion-fan> Hackman238: same errors as before...

[27-Sep-2011 12:18:29] <dhopp> it seems to be…but it would be nice to know wtf caused it

[27-Sep-2011 12:19:18] <ion-fan> Hackman238: one example: WARNING:zen.Events:Error processing transform/mapping on Event Class /

[27-Sep-2011 12:19:18] <ion-fan> Problem on line 2: AttributeError: 'NoneType' object has no attribute 'prepId'

[27-Sep-2011 12:19:18] <ion-fan> fs_id = device.prepId(evt.component)

[27-Sep-2011 12:20:04] <rocket> device is coming across as the value None

[27-Sep-2011 12:20:17] <rocket> you need to add a check ... if device:

[27-Sep-2011 12:20:32] <Hackman238> ion-fan: Very odd. I dont think this is a pack problem. Lets remove that zenpack again

[27-Sep-2011 12:20:35] <rocket> perhaps a trap or syslog event is coming in for a device that you are not monitoring yet

[27-Sep-2011 12:20:52] <rocket> that is a transform problem ..

[27-Sep-2011 12:21:07] <Hackman238> ion-fan: This time lets try a control

[27-Sep-2011 12:21:36] <Hackman238> ion-fan: Once you zenpack --remove that pack and rm the ZenPacks.SahenScott.DeviceSearch... directory

[27-Sep-2011 12:21:47] <Hackman238> ion-fan: Download servlet/JiveServlet/download/7453-7-2966/ZenPacks.zenoss.DeviceSearch-1.0.0-py2.6.egg.zip and unzip it

[27-Sep-2011 12:23:30] <ion-fan> done

[27-Sep-2011 12:23:34] <ion-fan> install?

[27-Sep-2011 12:23:37] <Hackman238> ion-fan: Okay zenpack --install that pack

[27-Sep-2011 12:23:44] <Hackman238> ion-fan: Yes please

[27-Sep-2011 12:24:45] <ion-fan> same problem

[27-Sep-2011 12:24:58] <Hackman238> ion-fan: Okay, rocket is correct, you'll need to fix the broken transform

[27-Sep-2011 12:25:19] <Hackman238> ion-fan: as suggested, you'll want to validate in the transform that device is not None.

[27-Sep-2011 12:27:53] <ion-fan> in my transform code?

[27-Sep-2011 12:27:57] <Hackman238> ion-fan: Yep

[27-Sep-2011 12:28:12] <ion-fan> i only pasted it from another forums-entry...

[27-Sep-2011 12:28:26] <ion-fan> after i looked over it qiuckly

[27-Sep-2011 12:28:30] <Hackman238> ion-fan: That adds to the likelyhood of it needing trweaking

[27-Sep-2011 12:28:37] <Hackman238> *tweaking

[27-Sep-2011 12:29:18] <ion-fan> my colleagues wanted 95th percentile so i pasted the code into that transform field...

[27-Sep-2011 12:29:21] <Hackman238> ion-fan: transforms usually need extra condition checking since they could be applied to a very large range of events

[27-Sep-2011 12:29:41] <Hackman238> ion-fan: Thats ok, but you'll want to add some extra condition checking.

[27-Sep-2011 12:29:46] <rocket> if your not checking for a device's existence before using it in a transform your transform will break on traps or syslogs

[27-Sep-2011 12:30:05] <Hackman238> ion-fan: Can you post the transform?

[27-Sep-2011 12:30:25] <ion-fan> i think its too long for irc

[27-Sep-2011 12:30:28] <ion-fan> email?

[27-Sep-2011 12:30:29] <rocket> how about the link to the transform so the thread can be updated

[27-Sep-2011 12:30:40] <Hackman238> rocket: Good idea

[27-Sep-2011 12:30:59] <Hackman238> ion-fan: Instead, can you post the link to the post?

[27-Sep-2011 12:33:39] <ion-fan> hackman238, rocket: message/52913#52913 that is the transform to drop naggy high-threshold-warnings on tunnel, dialer and virtual-access interfaces

[27-Sep-2011 12:33:46] <dhopp> If I have a device that will have no snmp monitoring so I don't want the default /Device template is it better to override it and then just make it a blank template or unbind it?  I don't want somebody to accidentally replace the binding with 'reset bindings'

[27-Sep-2011 12:35:31] <ion-fan> @hackman238: i mailed you my transform code

[27-Sep-2011 12:36:49] <JamesGamble> I'm having some trouble getting Zenoss to connect to WMI on a server, can anyone help?

[27-Sep-2011 12:36:54] <JamesGamble> I keep getting this error: Could not read the Windows event log (NT_STATUS_IO_TIMEOUT). Check your username/password settings and verify network connectivity.

[27-Sep-2011 12:37:12] <Hackman238> All, I need coffee. I'll BRB

[27-Sep-2011 12:37:21] <dhopp> JamesGamble:  do you have a firewall between zenoss and the windows machine?

[27-Sep-2011 12:37:27] <JamesGamble> I've updated the username and password on the server configuration, and the firewall is turned off

[27-Sep-2011 12:37:47] <JamesGamble> I even set the WMI timeout on the daemons (zeventlog and zwinwhatever) to 3 minutes

[27-Sep-2011 12:38:24] <JamesGamble> zeneventlog and zenwin actually

[27-Sep-2011 12:40:28] <ion-fan> @Hackman238: can we continue working tomorrow? because here it is 18:40 ^^

[27-Sep-2011 12:40:31] <JamesGamble> Additionally, the Zenoss server can ping the workstation

[27-Sep-2011 12:43:07] <Hackman238> ion-fan: Sure, I'm here most every day

[27-Sep-2011 12:43:17] <Hackman238> ion-fan: have a good night!

[27-Sep-2011 12:43:28] <Hackman238> ion-fan: Germany, right?

[27-Sep-2011 12:43:29] <ion-fan> ok talk you tomorrow...^^

[27-Sep-2011 12:43:32] <ion-fan> yes

[27-Sep-2011 12:43:39] <Hackman238> ion-fan: TTY then!

[27-Sep-2011 12:43:52] <ion-fan> im not going to sleep now only going at home...

[27-Sep-2011 12:44:01] <ion-fan> bye

[27-Sep-2011 12:44:27] <Hackman238> ion-fan: bye

[27-Sep-2011 12:45:09] <Hackman238> dhopp: either way its a one off.

[27-Sep-2011 12:45:26] <Hackman238> dhopp: I'd opt for the zProp one off though

[27-Sep-2011 12:45:34] <dhopp> but if I unbind the template wouldn't a reset bindings bring it back?

[27-Sep-2011 12:45:35] <tvincent_> whats the best way to figure out how many objects are being monitored?

[27-Sep-2011 12:45:39] <dhopp> oh ignore snmp

[27-Sep-2011 12:45:40] <dhopp> ?

[27-Sep-2011 12:45:44] <dhopp> that probably would be better

[27-Sep-2011 12:45:49] <Hackman238> dhopp: yeah

[27-Sep-2011 12:46:00] <Hackman238> tvincent_: http://yourzenoss.com:8080/manage

[27-Sep-2011 12:46:11] <Hackman238> tvincent_: Click on Control Panel, then Database

[27-Sep-2011 12:46:19] <Hackman238> tvincent_: Object count should be there somewhere

[27-Sep-2011 12:47:17] <Hackman238> tvincent_: http://yourzenoss.com:8080/Control_Panel/Database/main/manage_cacheParameters

[27-Sep-2011 12:47:32] <Hackman238> (direct to the page)

[27-Sep-2011 12:47:51] <Hackman238> JamesGamble: How are you supplying the username?

[27-Sep-2011 12:48:02] <tvincent_> thanks

[27-Sep-2011 12:48:09] <Hackman238> JamesGamble: as 'Domain\User', right?

[27-Sep-2011 12:48:12] <Hackman238> tvincent_: NP

[27-Sep-2011 12:48:13] <JamesGamble> Yep

[27-Sep-2011 12:48:21] <JamesGamble> In the "Configuration Properties" section

[27-Sep-2011 12:48:27] <JamesGamble> on the specific server that I want to monitor

[27-Sep-2011 12:49:09] <JamesGamble> Do I need to use the FDQN or no?

[27-Sep-2011 12:49:28] <JamesGamble> I've tried both, but I just want to know the correct format to use

[27-Sep-2011 12:49:37] <Hackman238> JamesGamble: Have you looked here? docs/DOC-4517

[27-Sep-2011 12:50:07] <JamesGamble> No, but the user I'm using is an admin user.

[27-Sep-2011 12:50:17] <Hackman238> JamesGamble: Ah ok.

[27-Sep-2011 12:51:30] <Hackman238> JamesGamble: From the zenoss box as zenoss user, can you try wmic -U [domain/]adminuser%password //host "select * from Win32_ComputerSystem" filling in domain, user and host?

[27-Sep-2011 12:52:46] <JamesGamble> it's a linux box

[27-Sep-2011 12:53:20] <Hackman238> JamesGamble: Yep, wmic should be in $ZENHOME/bin/wmic

[27-Sep-2011 12:53:44] <JamesGamble> alrighty, one sec

[27-Sep-2011 12:53:49] <JohnnyNOC> Hackman238 do you use zendisc to add devices?

[27-Sep-2011 12:53:54] <JohnnyNOC> wondering what's wrong with --prod_state

[27-Sep-2011 12:54:11] <JohnnyNOC> i can't seem to set them correctly..  tried for example "Development", "400"

[27-Sep-2011 12:54:41] <Hackman238> JohnnyNOC: I do, but I've never used that flag. I honestly dont know about it. I have a pack that sets all newly added devices to pre-production or any other prod state during the first model only

[27-Sep-2011 12:55:29] <JohnnyNOC> ah ok

[27-Sep-2011 12:55:29] <JamesGamble> how do i get to the $ZENHOME directory?

[27-Sep-2011 12:55:43] <JohnnyNOC> JamesGamble as the zenoss user, type cd and hit enter

[27-Sep-2011 12:55:43] <JohnnyNOC>

[27-Sep-2011 12:55:50] <JohnnyNOC> er

[27-Sep-2011 12:55:53] <JohnnyNOC> haha sorry i was confused

[27-Sep-2011 12:56:08] <JamesGamble> i don't login to the box as a zenoss user, just my normal account

[27-Sep-2011 12:56:19] <JohnnyNOC> it depends of course where zenoss was installed, but typically it's /usr/local/zenoss

[27-Sep-2011 12:56:28] <Hackman238> JamesGamble: Right but to test this we'll need to be the zenoss user

[27-Sep-2011 12:56:33] <JohnnyNOC> or is it /usr/local/zenoss/zenoss ?

[27-Sep-2011 12:57:22] <JamesGamble> wmic isn't at /usr/local/zenoss/zenoss/bin

[27-Sep-2011 12:57:46] <JohnnyNOC> JamesGamble can you use 'locate' to find it?

[27-Sep-2011 12:58:04] <JohnnyNOC> i believe this is what you want

[27-Sep-2011 12:58:07] <JohnnyNOC>  /usr/local/zenoss/common/bin/wmic

[27-Sep-2011 12:58:19] <JamesGamble> there it is

[27-Sep-2011 12:58:47] <rocket> as the zenoss user its best to use the convention $ZENHOME/bin etc

[27-Sep-2011 12:58:49] <Hackman238> JamesGamble: wmic is good for testing the pysamba stuff used for WMI and other Windows madness

[27-Sep-2011 12:59:01] <rocket> its valid between stack and rpm installs etc

[27-Sep-2011 12:59:41] <JamesGamble> running the command

[27-Sep-2011 13:00:06] <JohnnyNOC> rocket but that's not $ZENHOME/bin

[27-Sep-2011 13:00:13] <JamesGamble> it's taking awhile

[27-Sep-2011 13:00:22] <JohnnyNOC> i ran into that problem when setting up a custom datasource

[27-Sep-2011 13:00:27] <JamesGamble> nt status io timeout

[27-Sep-2011 13:00:34] <JohnnyNOC> had to symlink back to ../common/bin

[27-Sep-2011 13:00:40] <Hackman238> JamesGamble: okay it cant reach the target, probasbly need to open ports

[27-Sep-2011 13:00:41] <JohnnyNOC> probably wasn't the "right" thing to do

[27-Sep-2011 13:00:51] <JamesGamble> the firewall is off, and there's nothing in between

[27-Sep-2011 13:01:01] <Hackman238> JamesGamble: Hum. Windows firewall off also?

[27-Sep-2011 13:01:05] <JamesGamble> yes

[27-Sep-2011 13:01:13] <Hackman238> JamesGamble: WMI installed and the connectors setup?

[27-Sep-2011 13:01:14] <JohnnyNOC> oh, it was libexec i had to do that with

[27-Sep-2011 13:01:36] <JohnnyNOC> JamesGamble can you show us the command?

[27-Sep-2011 13:01:39] <Hackman238> JamesGamble: I can't recall how to do that off the top of my head. Let me see if I can find a doc

[27-Sep-2011 13:01:39] <rocket> JohnnyNOC: ok well you have done something nonstandard .. but my convention typically applies etc

[27-Sep-2011 13:01:39] <JamesGamble> sec

[27-Sep-2011 13:02:12] <JamesGamble> sudo ./wmic -U DOMAIN/Administrator%Password //X.X.X.X  "select * from Win32_ComputerSystem"

[27-Sep-2011 13:03:03] <Hackman238> JamesGamble: You'll defindatey want to run wmic as the zenoss user with no sudo

[27-Sep-2011 13:04:02] <JamesGamble> access denied when trying to login as zenoss

[27-Sep-2011 13:04:07] <JamesGamble> where is that password set?

[27-Sep-2011 13:04:22] <JamesGamble> logging into ssh as zenoss rather

[27-Sep-2011 13:05:00] <Hackman238> JamesGamble: You'll need to sudo su - or su root, then su zenoss

[27-Sep-2011 13:05:07] <JohnnyNOC> rocket i did that because i only wanted to type the scripts name in the data source

[27-Sep-2011 13:05:07] <JamesGamble> k

[27-Sep-2011 13:05:15] <JohnnyNOC> instead of like, ${here/libexec} or whatever it was supposed to be

[27-Sep-2011 13:05:42] <JohnnyNOC> but yes, i realized it was non-standard however i would like to set it up the "right" way

[27-Sep-2011 13:06:50] <JohnnyNOC> for example, my command template:

[27-Sep-2011 13:06:51] <JohnnyNOC> er

[27-Sep-2011 13:06:53] <JohnnyNOC> hah

[27-Sep-2011 13:06:57] <JohnnyNOC> check_mongodb.py -H ${here/manageIp} -P 27018 -D -A collections

[27-Sep-2011 13:07:20] <JamesGamble> it's asking me for the password to zenoss

[27-Sep-2011 13:07:25] <JohnnyNOC> but without seting up the symlink back to $ZENHOME/common/libexec it didn't work

[27-Sep-2011 13:07:44] <Hackman238> JamesGamble: i think the best thing to try is to verify WMI is replying using another Windows box to connect

[27-Sep-2011 13:07:44] <JamesGamble> nvm, i'm an idiot

[27-Sep-2011 13:07:49] <JamesGamble> i had the sudo in there

[27-Sep-2011 13:07:56] <Hackman238> JamesGamble:

[27-Sep-2011 13:07:59] <JamesGamble> i was able to verify from my own computer

[27-Sep-2011 13:08:14] <JamesGamble> waiting for the results of the wmic command as the zenoss user

[27-Sep-2011 13:08:39] <Hackman238> JamesGamble: Your computer sint the box we're trying to read WMI from, right?

[27-Sep-2011 13:08:45] <JamesGamble> got the timeout

[27-Sep-2011 13:09:02] <JamesGamble> Right

[27-Sep-2011 13:09:15] <JamesGamble> failed

[27-Sep-2011 13:09:17] <JamesGamble> io timeout

[27-Sep-2011 13:09:35] <JamesGamble> it's slow when i try on my laptop, but it does work

[27-Sep-2011 13:10:23] <JamesGamble> i tested on my xp laptop using the wbemtest utility

[27-Sep-2011 13:10:35] <JamesGamble> i should say i tested from my windows xp laptop to the server

[27-Sep-2011 13:12:24] <JohnnyNOC> JamesGamble i'm pretty sure you can increase the timeout for wmi if it's problematic

[27-Sep-2011 13:12:25] <Hackman238> JamesGamble: is the admin user local or domain?

[27-Sep-2011 13:12:33] <JamesGamble> I already have, to 3 minutes.

[27-Sep-2011 13:12:37] <JamesGamble> Either/or

[27-Sep-2011 13:12:40] <JamesGamble> I've tried both

[27-Sep-2011 13:13:05] <Hackman238> JamesGamble: Windows 2003 or 2008?

[27-Sep-2011 13:13:16] <JamesGamble> 2008 SP2 32-bit

[27-Sep-2011 13:15:32] <Hackman238> JamesGamble: try wmic -U 'yourdomain\youruser%password' //yourhost 'select * from win32_service'

[27-Sep-2011 13:15:42] <Hackman238> JamesGamble: Slightly differnt syntax

[27-Sep-2011 13:16:29] <JamesGamble> k, it's running now

[27-Sep-2011 13:17:26] <JamesGamble> What's the server default WMI timeout value?

[27-Sep-2011 13:17:44] <Hackman238> JamesGamble: Im not sure, I don't do a ton of WMI

[27-Sep-2011 13:18:05] <JamesGamble> I have it set to 3 minutes, but it's taking less than a minute for an error to come back

[27-Sep-2011 13:18:12] <Hackman238> JamesGamble: to remove the overriden value just delete that override on the properties page

[27-Sep-2011 13:18:37] <Hackman238> JamesGamble: Same error? IO Timeout?

[27-Sep-2011 13:18:47] <JamesGamble> yep, same error

[27-Sep-2011 13:19:41] <Hackman238> JamesGamble: Try the FQDN

[27-Sep-2011 13:19:43] <JamesGamble> [librpc/rpc/dcerpc_connect.c:790:dcerpc_pipe_connect_b_recv()] failed NT status (c00000b5) in dcerpc_pipe_connect_b_recv

[27-Sep-2011 13:19:43] <JamesGamble> [wmi/wmic.c:196:main()] ERROR: Login to remote object.

[27-Sep-2011 13:19:43] <JamesGamble> NTSTATUS: NT_STATUS_IO_TIMEOUT - NT_STATUS_IO_TIMEOUT

[27-Sep-2011 13:19:47] <Hackman238> JamesGamble: Less you;ve been using it

[27-Sep-2011 13:19:55] <JamesGamble> So DOMAIN.COM instead of DOMAIN?

[27-Sep-2011 13:21:00] <Hackman238> JamesGamble: Yes, host.domain.suffix

[27-Sep-2011 13:22:45] <JamesGamble> same thing

[27-Sep-2011 13:22:46] <JamesGamble> timeout

[27-Sep-2011 13:23:46] <Hackman238> JamesGamble: Not really sure. Seems like a reachability problem, but I'm not 100%. Generally WMI will give some version of NT_ACCESS_DENIED to debug

[27-Sep-2011 13:24:18] <JamesGamble> *sigh* ok

[27-Sep-2011 13:25:00] <Hackman238> JamesGamble: Maybe try to telnet to the remote port to see if it can connect?

[27-Sep-2011 13:25:25] <JamesGamble> I'll give that a go later... I need to grab some lunch

[27-Sep-2011 13:25:28] <JamesGamble> mental break ftw

[27-Sep-2011 13:26:14] <Hackman238> JamesGamble: Alrighty. We can ty to setup a fixed DCOM port and test connectivity when you get back

[27-Sep-2011 13:26:18] <Hackman238> JamesGamble: http://msdn.microsoft.com/en-us/library/windows/desktop/bb219447%28v=vs.85%29.aspx

[27-Sep-2011 13:34:36] <jbk> is there any particular process if i want to contribute a (minor) update to an existing community zenpack?

[27-Sep-2011 13:34:51] <jbk> (want to add some missing hw keys to the bluecoat zenpack)

[27-Sep-2011 13:35:12] <jmp242> Hmmm.

[27-Sep-2011 13:35:15] <jmp242> I think they'

[27-Sep-2011 13:35:24] <jmp242> re not currently in git yet, though they will be migrated

[27-Sep-2011 13:35:40] <jmp242> nyeates: any comment?

[27-Sep-2011 13:35:50] <jmp242> Nick is probably the guy currently

[27-Sep-2011 13:36:16] <jbk> is he in here?

[27-Sep-2011 13:37:40] <jmp242> yea, nyeates is the guy

[27-Sep-2011 13:37:45] <jmp242> but he's real busy today

[27-Sep-2011 13:37:47] <jbk> ahh  ok

[27-Sep-2011 13:37:50] <jmp242> so may afk

[27-Sep-2011 13:37:50] <jbk> no big rush

[27-Sep-2011 13:38:25] <jbk> just been playing around with it for a while, and am almost ready to really deploy it..

[27-Sep-2011 13:38:50] <jbk> i hope to also revamp the netapp zenpack..

[27-Sep-2011 13:39:06] <jbk> though the more i work with it, the more convinced i become that netapp doesn't get snmp

[27-Sep-2011 13:41:04] <jbk> (indexes in the filesystem table change whenever a volume is removed for example)

[27-Sep-2011 13:41:13] <jbk> or their snmp traps

[27-Sep-2011 13:43:46] <jmp242> mmm

[27-Sep-2011 13:43:49] <jmp242> Could be\

[27-Sep-2011 13:44:16] <jbk> the most annoying though is getting the snapshot usage data

[27-Sep-2011 13:44:27] <jbk> it's in the same table as the volumes

[27-Sep-2011 13:44:51] <jbk> basically for a volume w/ index 'x', the snapshot usage is an entry with index 'x+1'

[27-Sep-2011 13:45:04] <jbk> i'm still trying to figure out how to handle that in a template

[27-Sep-2011 13:45:44] <jbk> my inital guess is maybe a special-purpose data source

[27-Sep-2011 14:17:57] <JohnnyNOC> this guy Tito Ortega has become my daily lulz

[27-Sep-2011 14:23:53] <jb> finally got the new box for 4.

[27-Sep-2011 14:26:25] <Hackman238> JohnnyNOC: Whats he done now?

[27-Sep-2011 14:26:27] <Hackman238> jb: Nice

[27-Sep-2011 14:27:38] <JohnnyNOC> Hackman238 nothing really..  just reading his comments on the forums

[27-Sep-2011 14:27:46] <JohnnyNOC> one of them he signed himself as - Titonius

[27-Sep-2011 14:27:49] <JohnnyNOC> or something to that effect

[27-Sep-2011 14:28:03] <jb> ok, so 8 cores/12GB, RAID10 should be fine.

[27-Sep-2011 14:28:05] <jb> right?

[27-Sep-2011 14:28:20] <hmp> you call that new? :DD

[27-Sep-2011 14:28:41] <jb> ?

[27-Sep-2011 14:28:51] <jb> 8 cores is plenty.

[27-Sep-2011 14:28:54] <jb> otherwise I'd have more

[27-Sep-2011 14:29:17] <jb> and it's all I had laying around at the moment.

[27-Sep-2011 14:29:23] <Hackman238> jb: How much do you indend to monitor?

[27-Sep-2011 14:29:32] <jb> just 250-300 devices on this collector

[27-Sep-2011 14:29:40] <Hackman238> jb: More than enough

[27-Sep-2011 14:29:44] <Hackman238> jb: Should run sweet

[27-Sep-2011 14:29:49] <hmp> it was just a joke, I monitored 1000+ devices (hub + collector host) with that kind of machine

[27-Sep-2011 14:56:17] <Hackman238> hmp: Oh gotcha

[27-Sep-2011 14:56:48] <Jane_Curry> hiya

[27-Sep-2011 14:56:48] <jb> Hackman238: benefit to putting mysql on it's own filesystem?

[27-Sep-2011 14:57:46] <Jane_Curry> Does anyone have good examples / ZenPacks that provide dropdown menus?? I have one example where I provide menu items in objects.xml but I feel it should be possible to code these

[27-Sep-2011 14:57:47] <Hackman238> jb: I'd keep the sql db off the fs that hosts the rrd's

[27-Sep-2011 14:58:00] <Hackman238> jb: Impact in v3 wont be huge, but in v4 it would be

[27-Sep-2011 14:58:08] <jb> i wonder if this will screw me during the pgsql conversion

[27-Sep-2011 14:58:09] <jb> hm

[27-Sep-2011 14:58:12] <Hackman238> jb: You wont want relstore helped up by iowait during rrd commits

[27-Sep-2011 14:58:30] <Hackman238> s/helped/held/

[27-Sep-2011 14:58:38] <Hackman238> Jane_Curry: Hi Jane!

[27-Sep-2011 15:00:34] <jb> Hackman238: ok, so /opt/zenoss/Perf?

[27-Sep-2011 15:01:31] <Hackman238> jb: try to keep /opt/zenoss/perf seperated from whereever mysqld is

[27-Sep-2011 15:01:37] <jb> gotcha

[27-Sep-2011 15:02:02] <Hackman238> jb: Again, in v3 it matters not, but in v4 you definately dont want relstore helped up due to heavy rrd io

[27-Sep-2011 15:02:11] <Hackman238> s/helped/held/

[27-Sep-2011 15:02:16] <Hackman238> Sorry...long day

[27-Sep-2011 15:02:17] <Hackman238> LOL

[27-Sep-2011 15:02:20] <jb> yah

[27-Sep-2011 15:04:59] <jb> ok, i'm going to try and get LDAP/AD auth going from day 1 on this install.

[27-Sep-2011 15:05:56] <jb> Hackman238: hrm, so you wouldn't reccomed using a journaled filesystem for RRD data?

[27-Sep-2011 15:07:58] <pipeline> using a journaled FS shouldn't hurt a bit

[27-Sep-2011 15:08:26] <Hackman238> jb: journaling has some overhead, but its worth it.

[27-Sep-2011 15:08:41] <jb> so, just ext3?

[27-Sep-2011 15:08:50] <Hackman238> jb: I'd do ext3, yes

[27-Sep-2011 15:09:10] <jb>  Timing cached reads:   21724 MB in  2.00 seconds = 10881.00 MB/sec

[27-Sep-2011 15:09:10] <jb>  Timing buffered disk reads:  902 MB in  3.00 seconds = 300.64 MB/sec

[27-Sep-2011 15:09:17] <Hackman238> jb: have found ext4 to be slower with tiny tiny files.

[27-Sep-2011 15:09:49] <jb> guess thats not terrible.

[27-Sep-2011 15:10:02] <pipeline> a coworker told me he got better speeds out of the ext4 driver

[27-Sep-2011 15:10:04] <pipeline> reading ext3

[27-Sep-2011 15:10:10] <pipeline> than same filesystem using the ext3 driver

[27-Sep-2011 15:10:15] <pipeline> newer filesystems are broadly better, just in general

[27-Sep-2011 15:10:48] <pipeline> as far as gigantic fields of small files, it may be that the ext2/ext3/ext4 family is poorly suited.  Consider XFS.  (Unless you're on RHEL -- they charge extra for supported XFS )

[27-Sep-2011 15:11:53] <jb> yeah, i was thinking XFS

[27-Sep-2011 15:12:31] <Hackman238> jb: If xfs is an option, thats a great idea

[27-Sep-2011 15:12:47] <jb> it's EL5.

[27-Sep-2011 15:12:52] <jb> but, that doesn't mean it's not an option.

[27-Sep-2011 15:13:11] <jb> noatime/nodiratime on ext3 may be sufficient, though.

[27-Sep-2011 15:13:39] <Hackman238> jb: In general we use ext3, ext4 has horrible write throughput.

[27-Sep-2011 15:15:05] <jb> ARGH

[27-Sep-2011 15:15:09] <jb> vpn issues are killing me.

[27-Sep-2011 15:24:58] <dhopp> Hackman238:  I've tested ext4 and I'm getting good throughput….basically ext4 with ordered is getting same performance as ext3 with writeback

[27-Sep-2011 15:25:38] <Hackman238> jb: VPN always makes everything mroe fun!

[27-Sep-2011 15:25:52] <Hackman238> dhopp: Good to know, I dont think I tried that flag

[27-Sep-2011 15:26:25] <dhopp> mode=ordered is the default mount options….at least with the latest incarnation in RHEL 5.7

[27-Sep-2011 15:28:37] <Hackman238> dhopp: I doubt I tested since when it was very first available

[27-Sep-2011 15:34:43] <Hackman238> YAY! http://blog.timesunion.com/capitol/archives/82224/pef-members-shoot-down-contract-layoffs-to-commence/

[27-Sep-2011 15:36:28] <dhopp> Hackman238:  yeah I read that the early days of ext4 had horrible write performance

[27-Sep-2011 15:36:58] <Hackman238> dhopp: I'll have to retest soon

[27-Sep-2011 15:40:51] <Jane_Curry> Can anyone there comment on my request for help with creating dropdown menus @ 19:57 ?????????????????????

[27-Sep-2011 15:41:38] <Hackman238> Jane_Curry: Im sorry, I didnt see it

[27-Sep-2011 15:41:58] <Hackman238> Jane_Curry: I still dont see it. Can you repost?

[27-Sep-2011 15:42:58] <Jane_Curry> Does anyone have good examples / ZenPacks that provide dropdown menus?? I have one example where I provide menu items in objects.xml but I feel it should be possible to code these

[27-Sep-2011 15:43:28] <JohnnyNOC> Jane_Curry have you looked at the global search zenpack?

[27-Sep-2011 15:43:34] <JohnnyNOC> there's a dropdown, i guess you could consider it a menu

[27-Sep-2011 15:43:42] <JohnnyNOC> a dropdown of servers that it matches, which you can click on

[27-Sep-2011 15:43:47] <JohnnyNOC> perhaps there's something helpful in there?  :/

[27-Sep-2011 15:44:01] <Hackman238> Jane_Curry: Ah gotcha. To be honest the method depends on the menu

[27-Sep-2011 15:44:14] <Hackman238> Jane_Curry: My v3 ipsla pack has some menu foo in it

[27-Sep-2011 15:45:17] <Jane_Curry> Yeh - I know - I'll have a quick look at your IPSLA stuff

[27-Sep-2011 15:45:44] <Hackman238> Jane_Curry: Sorry, I wish I had much more to offer. The CloudFoundy pack and OpenStack packs do menus also

[27-Sep-2011 15:46:51] <Jane_Curry> What I am looking for is something akin to the Define Commands drop down off an Administration menu

[27-Sep-2011 15:47:55] <Jane_Curry> Hmm - wondererd about those 2 but I haven't found the dropdown menus yet - are they the Zenoss.Action bits???

[27-Sep-2011 15:50:35] <Hackman238> Jane_Curry: I'd have to look. I'll look after my meeting.

[27-Sep-2011 15:50:54] <Hackman238> Jane_Curry: Email me if I dont get back to you in a hour or so- I'm liable to forget.

[27-Sep-2011 16:00:10] <jb> hmm

[27-Sep-2011 16:00:13] <jb> so zends = mysql?

[27-Sep-2011 16:01:02] <pipeline> pretty much

[27-Sep-2011 16:01:40] <dhopp> how are people backing up their RRD files?

[27-Sep-2011 16:03:13] * JohnnyNOC raises his hand

[27-Sep-2011 16:03:40] <hmp> dhopp: i use bacula to pull the /perf directory (along with everything else)

[27-Sep-2011 16:04:51] * JohnnyNOC just rsyncs

[27-Sep-2011 16:07:51] <dhopp> JohnnyNOC: how often do you rsync?

[27-Sep-2011 16:07:55] <JohnnyNOC> nightly

[27-Sep-2011 16:08:00] <dhopp> and how many RRD files bacula/JohnnyNOC ?

[27-Sep-2011 16:08:04] <dhopp> err

[27-Sep-2011 16:08:07] <dhopp> hmp

[27-Sep-2011 16:09:37] <JohnnyNOC> i don't know let me count

[27-Sep-2011 16:09:37] <JohnnyNOC>

[27-Sep-2011 16:09:51] <JohnnyNOC> ~1000 devices on 3 collectors

[27-Sep-2011 16:10:34] <dhopp> JohnnyNOC: find /opt/zenoss/perf -name *.rrd | wc -l

[27-Sep-2011 16:10:35] <dhopp> :-P

[27-Sep-2011 16:10:53] <JohnnyNOC> *nod*

[27-Sep-2011 16:10:55] <JohnnyNOC> waiting

[27-Sep-2011 16:10:56] <hmp> around 67k

[27-Sep-2011 16:11:03] <JohnnyNOC> 312k

[27-Sep-2011 16:12:25] <dhopp> JohnnyNOC: so ~100K per collector?  how long does the rsync take?

[27-Sep-2011 16:12:33] <dhopp> hmp: about how long does the backup take?

[27-Sep-2011 16:12:43] <JohnnyNOC> dhopp no they're spread out unevenly based upon their location

[27-Sep-2011 16:12:44] <dhopp> I'm trying to decide best way to do it

[27-Sep-2011 16:12:49] <dhopp> gotcha

[27-Sep-2011 16:13:12] <dhopp> part of me thinks rsync to another server and then backup that server with a backup agent

[27-Sep-2011 16:13:32] <JohnnyNOC> rsync to san

[27-Sep-2011 16:13:34] <JohnnyNOC>

[27-Sep-2011 16:13:45] <dhopp> JohnnyNOC: need a SAN first :-P

[27-Sep-2011 16:13:56] <JohnnyNOC> heheh understood

[27-Sep-2011 16:14:15] <JohnnyNOC> that would be great of course, but you wouldn't be comfortable with 2 copies of yoru RRDs?

[27-Sep-2011 16:14:16] <JohnnyNOC> you need 3?

[27-Sep-2011 16:14:17] <JohnnyNOC>

[27-Sep-2011 16:17:34] <jmp242> hey anyone using an EL (RHEL, CENTOS,SL)5.7 and now getting filesystem alerts on component cached memory when they didn't before?

[27-Sep-2011 16:17:45] <jmp242> is this something I can just add to zfilesystemignoretypes?

[27-Sep-2011 16:18:13] <jmp242> i.e on 5.6

[27-Sep-2011 16:20:40] <JohnnyNOC> jmp242 i believe someone was in here with the same type of issue and said they added it to zfilesystemignoretypes

[27-Sep-2011 16:21:16] <jmp242> yea that's what I thought

[27-Sep-2011 16:24:38] <Hackman238> dhopp: copy from fusion IO RAID10 to spindle RAID10 every 30 minutes

[27-Sep-2011 16:26:09] <JohnnyNOC> <3 fusion io

[27-Sep-2011 16:26:19] <JohnnyNOC> Hackman238 why don't you have a zenpack for it?

[27-Sep-2011 16:26:19] <JohnnyNOC>

[27-Sep-2011 16:26:31] <JohnnyNOC> i put together a basic monitoring template and have been considering packaging it up for release

[27-Sep-2011 16:26:42] <jmp242> Hackman238: or JohnnyNOC -

[27-Sep-2011 16:26:51] <jmp242> do you happen to know the "name" Zenoss is using

[27-Sep-2011 16:27:05] <jmp242> for the cached memory and memory buffers

[27-Sep-2011 16:27:07] <Hackman238> [zenoss@zen5.dfw1 ~]$ find /opt/zenoss/perf -name *.rrd | wc -l

[27-Sep-2011 16:27:07] <Hackman238> 308170

[27-Sep-2011 16:27:15] <Hackman238> Thats just Zen5 in DFW

[27-Sep-2011 16:27:18] <Hackman238>

[27-Sep-2011 16:27:19] <jmp242> so I can put them in zproperty lll

[27-Sep-2011 16:27:28] <jmp242> for the ignore

[27-Sep-2011 16:27:40] <JohnnyNOC> jmp242 i have no idea..  we've been moving away from centos

[27-Sep-2011 16:27:48] <JohnnyNOC> there's a channel logger

[27-Sep-2011 16:27:58] <JohnnyNOC> maybe the convo from earlier today is online somewhere?

[27-Sep-2011 16:28:13] <JohnnyNOC> otherwise, don't you just know what the name is by looking at it?  i must not get it

[27-Sep-2011 16:28:17] <Hackman238> [zenoss@zen4.dfw1 ~]$ find /opt/zenoss/perf -name *.rrd | wc -l

[27-Sep-2011 16:28:18] <Hackman238> 623012

[27-Sep-2011 16:28:29] <jmp242> I'll look

[27-Sep-2011 16:28:48] <Hackman238> jmp242: I dont know what you mean

[27-Sep-2011 16:29:15] <dhopp> Hackman238:  do you have a fuison IO RAID 10 I can have?

[27-Sep-2011 16:29:17] <dhopp> didn't think so

[27-Sep-2011 16:29:24] <jmp242> so I have filesystem full errors in SL5.7

[27-Sep-2011 16:29:34] <jmp242> for component cached memory and component memory buffers

[27-Sep-2011 16:29:37] <JohnnyNOC> dhopp $9k

[27-Sep-2011 16:29:38] <JohnnyNOC> cash

[27-Sep-2011 16:29:39] <jmp242> it sounds like RAM stuff

[27-Sep-2011 16:29:41] <JohnnyNOC> pickup in the loop

[27-Sep-2011 16:29:42] <JohnnyNOC>

[27-Sep-2011 16:30:02] <jmp242> I feel like I should be able to do a zfilesystemignoretype entry

[27-Sep-2011 16:30:09] <JohnnyNOC> i had 2 in my car once and as i was driving to the DC i felt poor because they were worth more than my car

[27-Sep-2011 16:30:10] <JohnnyNOC> :/

[27-Sep-2011 16:30:53] <JohnnyNOC> 1 of them is more expensive than my car!  hah

[27-Sep-2011 16:33:13] <Hackman238> dhopp: Unfortunately not LOL

[27-Sep-2011 16:33:24] <Hackman238> dhopp: So much $$ its crazy

[27-Sep-2011 16:33:36] <Hackman238> Almost cheaper for us to build a private Zenoss SAN

[27-Sep-2011 16:33:55] <JohnnyNOC> yea but it wouldn't be as fast

[27-Sep-2011 16:34:08] <JohnnyNOC> unless maybe you're talking about a san full or ssds

[27-Sep-2011 16:34:12] <Hackman238> JohnnyNOC: Not even close

[27-Sep-2011 16:34:14] <Hackman238> XD

[27-Sep-2011 16:34:16] <JohnnyNOC> even then..  the fusion io is still faster

[27-Sep-2011 16:34:23] <JohnnyNOC> hehehe

[27-Sep-2011 16:34:30] <JohnnyNOC> our dbas LOVE our fusion-ios to death

[27-Sep-2011 16:34:42] <dhopp> Hackman238:  I'm assuming your DFW boxes are broken up into logical collectors...

[27-Sep-2011 16:34:48] <Hackman238> JohnnyNOC: What sort of car do you have?

[27-Sep-2011 16:34:55] <JohnnyNOC> jeep cherokee

[27-Sep-2011 16:34:56] <JohnnyNOC>

[27-Sep-2011 16:34:58] <JohnnyNOC> 2001

[27-Sep-2011 16:35:04] <JohnnyNOC> limited, 4x34

[27-Sep-2011 16:35:05] <Hackman238> dhopp: 2 to 4 logical collectors on each physical collector

[27-Sep-2011 16:35:06] <JohnnyNOC> 4x4

[27-Sep-2011 16:35:13] <jb> well, i see the LDAP stuff is still buggy

[27-Sep-2011 16:35:14] <jb> ugh.

[27-Sep-2011 16:35:16] <JohnnyNOC> probably worth $4k in a private sale

[27-Sep-2011 16:35:20] <JohnnyNOC> jb i don't have problems with it

[27-Sep-2011 16:35:30] <Hackman238> dhopp: up to 6 physcial collectors excluding the master and db server

[27-Sep-2011 16:35:34] <jb> hrm one sec

[27-Sep-2011 16:35:38] <Hackman238> JohnnyNOC: Yikes.

[27-Sep-2011 16:35:44] <JohnnyNOC> jb whats the problem?

[27-Sep-2011 16:35:59] <Hackman238> jb: Yeah...unfortunately

[27-Sep-2011 16:35:59] <JohnnyNOC> Hackman238 i love my jeep

[27-Sep-2011 16:36:08] <JohnnyNOC> i'll drive it til it's worth $1 doesn't matter to me really

[27-Sep-2011 16:36:28] <jb> just setting it up now.

[27-Sep-2011 16:36:32] <JohnnyNOC> perfect for the city

[27-Sep-2011 16:36:45] <jb> but, it is stupid that it authenticates every click.

[27-Sep-2011 16:36:46] <JohnnyNOC> big enough to carry shit around, small enough to fit easily into parking spots

[27-Sep-2011 16:37:00] <jb> JohnnyNOC: are you using it wth AD?

[27-Sep-2011 16:37:19] <JohnnyNOC> my jeep?  no.

[27-Sep-2011 16:37:22] <JohnnyNOC> zenoss?  yes.

[27-Sep-2011 16:37:23] <JohnnyNOC>

[27-Sep-2011 16:37:31] <jb>

[27-Sep-2011 16:37:40] <jb> um, do you pre-create accounts and assign roles in zenoss?

[27-Sep-2011 16:38:11] <jmp242> Hey, is there an undo lol for deleting a local zproperty? Or somewhere I can find the old value/

[27-Sep-2011 16:38:11] <jmp242> ?

[27-Sep-2011 16:38:53] <dhopp> jmp242:  doesn't it say "there is no undo for this" when you tell it to delete a local copy?

[27-Sep-2011 16:38:54] <dhopp> :-P

[27-Sep-2011 16:39:03] <jmp242> No, but I know that

[27-Sep-2011 16:39:09] <jmp242> it's a stupid text string lol

[27-Sep-2011 16:39:20] <Hackman238> JohnnyNOC: I know how you mean, I primarily drive my Civic

[27-Sep-2011 16:40:08] <Hackman238> jb: Thats what we do...

[27-Sep-2011 16:40:27] <Hackman238> jb: Unfortunately its the only way it seems to take corectly

[27-Sep-2011 16:40:40] <jb> Ok, thats good for me.

[27-Sep-2011 16:40:56] <jb> I just logged in with an AD user, and I don't see any devices.

[27-Sep-2011 16:40:58] <jb> which is good.

[27-Sep-2011 16:41:06] <jb> because I don't want just any AD user to access zenoss.

[27-Sep-2011 16:41:19] <Hackman238> jb: Good idea.

[27-Sep-2011 16:41:34] <jb> erm

[27-Sep-2011 16:41:34] <jb> 2011-09-27T15:41:21 INFO event.LDAPUserFolder login failed: admin (no such user)

[27-Sep-2011 16:41:38] <jb> see, grr.

[27-Sep-2011 16:41:42] <jb> now I can't use local users?

[27-Sep-2011 16:42:12] <Hackman238> jb: You should be able to

[27-Sep-2011 16:42:37] <jb> hrm, apparently not.

[27-Sep-2011 16:42:46] <jb> admin isn't working.

[27-Sep-2011 16:43:19] <jmp242> ah ha!

[27-Sep-2011 16:43:23] <jmp242> zope Undo works YAY

[27-Sep-2011 16:43:52] <Hackman238> jb: thats odd.

[27-Sep-2011 16:44:49] <Hackman238> jb: Its working here. v4.0.2 with LDAP auth version 2.0.5, right?

[27-Sep-2011 16:45:01] <JohnnyNOC> jb no i don't pre-create anything

[27-Sep-2011 16:45:17] <JohnnyNOC> i dont even think i messed with the role mapping too much

[27-Sep-2011 16:45:42] <jb> yep

[27-Sep-2011 16:46:03] <jb> so, you have the new users login, which makes them appear in "USERS," then you set their roles?

[27-Sep-2011 16:50:11] <JohnnyNOC> i have some mappings setup

[27-Sep-2011 16:50:12] <JohnnyNOC> user

[27-Sep-2011 16:50:14] <JohnnyNOC> zenmanager

[27-Sep-2011 16:50:17] <JohnnyNOC> zenowner i think?

[27-Sep-2011 16:50:22] <JohnnyNOC> 1 allows them to do nothing

[27-Sep-2011 16:50:27] <JohnnyNOC> 1 allows them to manage a device

[27-Sep-2011 16:50:31] <JohnnyNOC> another gives them accss to zope

[27-Sep-2011 16:50:36] <JohnnyNOC> (if i recall correctly)

[27-Sep-2011 16:51:02] <Hackman238> JohnnyNOC: manager, zenmanager, zenuser

[27-Sep-2011 16:51:17] <Hackman238> JohnnyNOC: And its "lord of all that is Zope"

[27-Sep-2011 16:51:20] <Hackman238>

[27-Sep-2011 16:51:24] <JohnnyNOC> hahaha

[27-Sep-2011 16:56:43] <JohnnyNOC> that person is me

[27-Sep-2011 16:56:45] <JohnnyNOC> :-P

[27-Sep-2011 16:57:21] <JohnnyNOC> oh, and a few other people too..  i need to fix my mappings

[27-Sep-2011 16:57:24] <JohnnyNOC> :/

[27-Sep-2011 16:57:33] <Hackman238> JohnnyNOC: LOL

[27-Sep-2011 16:58:25] <JohnnyNOC> hah, damn

[27-Sep-2011 16:58:28] <JohnnyNOC> failboat

[27-Sep-2011 16:58:33] <Hackman238> JohnnyNOC: Not just anyone can juggle the chainsaws of power and authority commanded by a zope user!

[27-Sep-2011 16:58:51] <JohnnyNOC> absolutely

[27-Sep-2011 16:59:02] <JohnnyNOC> if one of these guys fucks me i'll be pissed!

[27-Sep-2011 16:59:33] <Hackman238> JohnnyNOC: I'd be too- that'll hurt

[27-Sep-2011 16:59:59] <JohnnyNOC> i need Shane Scott on speed dial

[27-Sep-2011 17:00:08] <dhopp> argh..I know that there was a transformation that somebody recommended to keep the mysql db from exploding with events, but I can't seem to find it…I think it dropped all the set events

[27-Sep-2011 17:00:42] <JohnnyNOC> dhopp you mean if like, the device is pinging down already?

[27-Sep-2011 17:00:47] <jb> so what are you guys monitoring by default (graph/perf wise) on all linux boxes?

[27-Sep-2011 17:00:55] <jb> anything other than what zenoss ships with?

[27-Sep-2011 17:01:20] <JohnnyNOC> i monitor ssswapin/ssswapout

[27-Sep-2011 17:01:28] <JohnnyNOC> also nic discards/drops

[27-Sep-2011 17:02:05] <Hackman238> /Events/Change/Set -> action:drop

[27-Sep-2011 17:02:11] <JohnnyNOC> i'd certainly be interested in hearing what otherws like Hackman238 are doing

[27-Sep-2011 17:02:12] <dhopp> JohnnyNOC: no I meant all of the "set attribute 'port' to '871' on object 'udp_00871'" messages

[27-Sep-2011 17:02:52] <jb> yeah def

[27-Sep-2011 17:02:55] <jb> what about you hackman?

[27-Sep-2011 17:03:01] <JohnnyNOC> dhopp ah sorry, i'm not familiar with those

[27-Sep-2011 17:03:05] <Hackman238> What do I monitor?

[27-Sep-2011 17:03:11] <JohnnyNOC> mysql monitoring is on my todo list

[27-Sep-2011 17:03:20] <JohnnyNOC> but we use postgres/edb/msssql more

[27-Sep-2011 17:03:44] <Hackman238> Just network gear and Zenoss. ~140K devices at this point

[27-Sep-2011 17:03:49] <JohnnyNOC> Hackman238 what else besides the default are you monitoring and graphing on your linux hosts

[27-Sep-2011 17:04:08] <Hackman238> JohnnyNOC: Load, memory free. nothing more.

[27-Sep-2011 17:04:25] <dhopp> Hackman238:  how many people make changes to Zenoss (i.e. add/remove devices/templates)?

[27-Sep-2011 17:05:31] <jb> Hackman238: oh, why so minimal?

[27-Sep-2011 17:05:42] <Hackman238> dhopp: Here? Just my team, mostly me. Everything is added automagically using a custom zendisc setup and our ticketing systems

[27-Sep-2011 17:05:53] <JohnnyNOC> jb i think because they ahve so many servers

[27-Sep-2011 17:05:56] <JohnnyNOC> they = rackspace

[27-Sep-2011 17:06:04] <JohnnyNOC> but i'll let him answer that one

[27-Sep-2011 17:06:11] <dhopp> Hackman238:  that's what I thought...

[27-Sep-2011 17:06:16] <Hackman238> jb: theres nothing else we need to monitor besides the zenoss processes

[27-Sep-2011 17:06:36] <Hackman238> We add about 250 devices a day

[27-Sep-2011 17:07:24] <Hackman238> Well about that many automatically

[27-Sep-2011 17:07:38] <Hackman238> Sometimes up to several hundred.

[27-Sep-2011 17:07:48] <Hackman238> Its really sort of scarey we're running v2.5.2

[27-Sep-2011 17:08:25] <Hackman238>

[27-Sep-2011 17:08:38] <Hackman238> I have to run

[27-Sep-2011 17:08:41] <Hackman238> Later all!

[27-Sep-2011 17:46:36] <jb> hrm, so I updated my modeler plugins for /Server, but when I model a device, they aren't getting included.

[27-Sep-2011 17:46:41] <jb> restarted zenoss just in case.

[27-Sep-2011 17:47:13] <jb> ah, they didn't get inherited by /Server/Linux

[27-Sep-2011 17:47:23] <jb> is there a setting for inheritence now?

[27-Sep-2011 17:47:42] <dhopp> jb: if the property is defined at /Server/Linux it will override what is above it

[27-Sep-2011 19:31:18] <orudie> which language is zenoos written in ?

[27-Sep-2011 19:31:23] <orudie> errr

[27-Sep-2011 19:31:27] <orudie> zenoss*

[27-Sep-2011 19:40:14] <orudie> anyone ?

[27-Sep-2011 20:09:36] <nexex> orudie: python mostly

[27-Sep-2011 21:03:32] <rocket> btw there is a zproperty the prevents the change events from being logged at all

[27-Sep-2011 21:03:43] <rocket> no need to even drop them

[27-Sep-2011 21:04:43] <rocket> orudie: python and java

[27-Sep-2011 21:05:02] <rocket> java is more in the enterprise edition, there is some c in use for wmi bits

[28-Sep-2011 00:00:01] [disconnected at Wed Sep 28 00:00:01 2011]

[28-Sep-2011 00:00:02] [connected at Wed Sep 28 00:00:02 2011]

[28-Sep-2011 00:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[28-Sep-2011 04:06:14] <battaglin> hi

[28-Sep-2011 04:06:29] <ion-fan> hackman238: im ready for debugging..^^

[28-Sep-2011 04:06:54] <battaglin> im missing my windows services under the Components tab on one server view.. what am I missing??

[28-Sep-2011 06:23:03] battaglin is now known as MaDSeN

[28-Sep-2011 06:23:15] MaDSeN is now known as battaglin

[28-Sep-2011 08:24:53] <jmp242> Morning all

[28-Sep-2011 08:40:29] <jb> morning.

[28-Sep-2011 08:55:09] <JohnnyNOC> good morning everybody

[28-Sep-2011 08:55:30] <JohnnyNOC> </Dr. Nick Rivera>

[28-Sep-2011 08:55:39] <jb> small UI bug in 4.. when you add a location, the window doesn't refresh to show the new location.

[28-Sep-2011 08:55:59] <JohnnyNOC> jb i haven't seen the UI in 4..  how is it?

[28-Sep-2011 08:56:06] <JohnnyNOC> will i see a screenshot on zenoss.com somewhere

[28-Sep-2011 08:56:42] <jb> just started using it

[28-Sep-2011 08:56:46] <jb> its very similar to 3

[28-Sep-2011 08:56:52] <jb> however, all of the crappy ajax stuff appears to be fixed

[28-Sep-2011 08:57:04] <jb> small little tweaks here and there

[28-Sep-2011 08:57:29] <JohnnyNOC> nice

[28-Sep-2011 09:04:39] <jb> ugh.

[28-Sep-2011 09:05:01] <jb> zope/zenhub/zenjobs suddenly wont start now.

[28-Sep-2011 09:05:15] <Hackman238> jb: make sure zeneventserver is running

[28-Sep-2011 09:05:22] <Hackman238> jb: Also that zends is running

[28-Sep-2011 09:05:37] <Hackman238> jb: And that memcached and rabbitmq-server haven't bombed

[28-Sep-2011 09:05:52] <jb> ah, zends wasn't running.

[28-Sep-2011 09:05:54] <jb> that was the problem.

[28-Sep-2011 09:05:56] <Hackman238> jb:

[28-Sep-2011 09:06:04] <jb> i just discovered that as you typed it

[28-Sep-2011 09:06:17] <jb> morning, btw.

[28-Sep-2011 09:06:41] <Hackman238> jb: Morning

[28-Sep-2011 09:06:44] <Hackman238> Morning all!

[28-Sep-2011 09:07:19] <jb> i need to play with this LDAP stuff some more.. i need to be able to pre-assign roles/acl's to LDAP users, which means I'll have to pre-create them before they login.  hopefully that works.

[28-Sep-2011 09:11:16] <zirikili> hi all

[28-Sep-2011 09:11:37] <jb> hi

[28-Sep-2011 09:14:12] <zirikili> I have a question about zenoss.  I have some device cataloged on /Ping class. some of them does not repond on ping but zenoss does not list it as a problem. what could be wrong?

[28-Sep-2011 09:15:50] <Hackman238> jb: Yeah LDAP and ACL's have always been tricky, but they can be accomplished

[28-Sep-2011 09:16:12] <Hackman238> zirikili: You have a device in /Ping that is not pining but its not flagged as down?

[28-Sep-2011 09:16:42] <zirikili> Hackman238:  yes… exactly

[28-Sep-2011 09:17:26] <Hackman238> zirikili: Ensure that the device doesnt have zPingMonitorIgnore set to True and that the device isn't in Maint. production state. If this is so, navigate to the device status page and click the gear in the lower left -> push changes. It could be the zenping daemon isnt testing for this device.

[28-Sep-2011 09:18:11] <zirikili> Hackman238:  hmmm… let me check

[28-Sep-2011 09:22:33] <jb> odd, I can't auth to zends with a user that I created.

[28-Sep-2011 09:32:07] <Hackman238> jb: did you grant yourself rights?

[28-Sep-2011 09:32:22] <Hackman238> zirikili: Hows that look?

[28-Sep-2011 09:32:42] <zirikili> Hackman238: Just a minute

[28-Sep-2011 09:33:06] <jb> Hackman238: just trying to create a zends user so I can monitor it with MysqlMonitor.. a "create user username identified by 'blah'" should be sufficient for this, no?

[28-Sep-2011 09:33:55] <Hackman238> jb: I honestly dont know...thats a google question.

[28-Sep-2011 09:34:06] <Hackman238> jb: I'd say try granting super and see if that works

[28-Sep-2011 09:34:19] <jb> alrighty

[28-Sep-2011 09:34:29] <Hackman238> zirikili: Alrighty

[28-Sep-2011 09:37:32] <rocket> jb other than running on a different port and socket zends is really just mysql ... oracle wouldnt let us call it that though

[28-Sep-2011 09:37:52] <jb> yeah i noticed it's the enterprise edition?

[28-Sep-2011 09:38:14] <rocket> jb: I am not sure which edition it is.. I think it is though

[28-Sep-2011 09:39:22] <Hackman238> rocket: Does one need super on a user to use the mysqlmonitor?

[28-Sep-2011 09:40:01] <zirikili> Hackman238:  ZenPingMonitorIgnore is False

[28-Sep-2011 09:40:02] <rocket> Hackman238: I dont recall needing that ..

[28-Sep-2011 09:40:14] <Hackman238> zirikili: Okay good

[28-Sep-2011 09:40:27] <rocket> zirikili: what version of zenoss?

[28-Sep-2011 09:40:37] <zirikili> Hackman238: Production State is Production

[28-Sep-2011 09:40:53] <Hackman238> zirikili: Alrighty. I'd navigate to the device status page and push changes.

[28-Sep-2011 09:41:09] <Hackman238> zirikili: Just to make sure zenping was notified

[28-Sep-2011 09:42:06] <pipeline> heh so zenoss enterprise now starts at $37,500 http://www.zenoss.com/solution/pricing

[28-Sep-2011 09:42:23] <rocket> if its a version of zenoss before 3.2 I would just restart zenping .. zenping most likely has gotten stuck waiting for device config changes

[28-Sep-2011 09:43:20] <ion-fan> @Hackman238: do you have more ideas what to do?

[28-Sep-2011 09:43:39] <rocket> pipeline: and what about competitors charging for their monitoring tool?

[28-Sep-2011 09:44:15] <Hackman238> rocket: That was going to be my next suggestion

[28-Sep-2011 09:44:21] <Hackman238> rocket: BTW, Morning

[28-Sep-2011 09:44:27] <zirikili> Hackman238: Where do I cheek zenoss version ?

[28-Sep-2011 09:44:37] <Hackman238> ion-fan: I forget the problem, can refresh my memory?

[28-Sep-2011 09:44:45] <pipeline> rocket: depends what you think their competitors are

[28-Sep-2011 09:44:48] <Hackman238> zirikili: su root; rpm-qa zenoss

[28-Sep-2011 09:44:56] <ion-fan> message/61582#61582

[28-Sep-2011 09:45:05] <Hackman238> *su root; rpm -qa zenoss

[28-Sep-2011 09:45:07] <ion-fan> problem with your zenpack

[28-Sep-2011 09:45:15] <pipeline> rocket: they've clearly decided they compete with the likes of CA and HP, and I'm not entirely sure that's the kind of vendor I want to deal with :X

[28-Sep-2011 09:45:32] <Hackman238> ion-fan: Oh right right. I posted a change for the transform

[28-Sep-2011 09:45:34] <zirikili> Hackman238:  I am using Debian… but I now know how to check

[28-Sep-2011 09:45:47] <rocket> pipeline: to be in full disclosure I am a zenoss employee

[28-Sep-2011 09:46:03] <pipeline> rocket: well kudos on community engagement

[28-Sep-2011 09:46:43] <Hackman238> ion-fan: give that change a wirl and restart zenoss. Then try installing the pack again.

[28-Sep-2011 09:47:00] <rocket> I dont understand why competing with certain companies gives you pause as to what kind of company zenoss is

[28-Sep-2011 09:50:47] <rocket> pipeline: it sounds like you have concerns .. well we work very hard as a company to meet customer support requirements etc.  We do have a fast moving innovative product on the market and I think that deserves getting paid for it.

[28-Sep-2011 09:51:20] <zirikili> Hackman238: the version is 3.2.0

[28-Sep-2011 09:52:09] <pipeline> rocket: steep price increase; zenoss 4 does not appear to have a 1:1 relationship with the open source edition.  Disturbing events.

[28-Sep-2011 09:52:14] <Hackman238> ion-fan: Sorry, I updated my post because of a logic error

[28-Sep-2011 09:53:00] <rocket> pipeline: the 1:1 relationship is returning as far as I know with zenoss 4.2

[28-Sep-2011 09:53:04] <Hackman238> zirikili: I'd try to push the change and give it two minutes

[28-Sep-2011 09:53:15] <Hackman238> zirikili: If it doesnt work, su zenoss; zenping restart

[28-Sep-2011 09:53:53] <rocket> honestly I dont know what the price was before but I dont think it has been a steep price increase.

[28-Sep-2011 09:53:59] <pipeline> rocket: that is actually hugely reassuring.   I saw material from Zenoss suggesting that there wouldn't be a 4.x open source release, but more releases of 3.x instead.  I must have misinterpreted it.

[28-Sep-2011 09:54:03] <Hackman238> pipeline: Zenoss support does an excellent job with us here at Rackspace

[28-Sep-2011 09:54:29] <dhopp> Hackman238:  How often do you call support?

[28-Sep-2011 09:54:43] <rocket> hourly!

[28-Sep-2011 09:54:55] <Hackman238> dhopp: I put in tickets when I can't figure something out

[28-Sep-2011 09:54:57] <rocket> just kidding

[28-Sep-2011 09:55:08] <Hackman238> dhopp: I try not to call less I have to...I hate the phone.

[28-Sep-2011 09:55:11] <Hackman238> LOL

[28-Sep-2011 09:55:25] <rocket> most of our cases come in through our portal ..

[28-Sep-2011 09:55:40] <rocket> and we do get some doozies to figure out ..

[28-Sep-2011 09:56:20] <rocket> it often times ends up figuring out the quirk with the remotely monitored device.  For some reason other vendors dont like every standard ..

[28-Sep-2011 09:58:43] <Hackman238> rocket: There's only 16 standards!

[28-Sep-2011 09:58:55] <rocket> only

[28-Sep-2011 09:59:26] <Hackman238> rocket: It kills me that the idea of standardization conteracts itself by creating more than one standard...

[28-Sep-2011 10:00:09] <SDuensin> Morning.

[28-Sep-2011 10:00:24] <rocket> my view is that support must be doing something correctly if we can make Rackspace think we have excellent support when their motto is "fanatical support"

[28-Sep-2011 10:00:59] <SDuensin> Um, some fanatics blow things up.

[28-Sep-2011 10:01:14] <rocket> shush!

[28-Sep-2011 10:01:21] <SDuensin> hehe

[28-Sep-2011 10:01:49] <Hackman238> SDuensin: LOL!

[28-Sep-2011 10:01:50] <dhopp> Hackman238:  Windows has 16 standards…depending on what version of Windows you are using

[28-Sep-2011 10:02:01] <Hackman238> dhopp: LOL

[28-Sep-2011 10:04:40] <Hackman238> ion-fan: Let me know if that re-posted code works for you

[28-Sep-2011 10:09:07] <ion-fan> ok i let you know sorry i have to do some other things before testing that...

[28-Sep-2011 10:10:02] <Hackman238> ion-fan: No problem. I'll be here.

[28-Sep-2011 10:21:35] <zirikili> Hackman238:  I will try upgrade the zenoss pack… be right back…. thank you

[28-Sep-2011 10:22:39] <Hackman238> zirikili: ok

[28-Sep-2011 10:49:48] <ion-fan> Hackman238: YEAAHH no errors!!

[28-Sep-2011 10:50:24] <Hackman238> ion-fan: Very cool

[28-Sep-2011 10:50:43] <ion-fan> Hackman238: can you tell me again what i have to insert in the__init__.py-file?

[28-Sep-2011 10:52:20] <Hackman238> ion-fan: I'll post it up in a bit, I have to figure it out. I dont think the code I gave you before will work.

[28-Sep-2011 10:52:52] <ion-fan> Hackman238: for installation i used your modified .egg file... the row numbers seem to be different..

[28-Sep-2011 10:53:12] <ion-fan> ok

[28-Sep-2011 10:53:24] <Hackman238> ion-fan: They might be, I shortened a line pretty significantly

[28-Sep-2011 10:53:41] <Hackman238> I'll have to post an update in a day or so

[28-Sep-2011 10:54:36] <jb> hrm, ive got to figure out how to get bind 9.7 style stats into zen.

[28-Sep-2011 10:54:43] <jb> since my old method no longer works with newer versions of bind :/

[28-Sep-2011 11:01:41] <ion-fan> Hackman238: so i have to wait till you fixed the code?

[28-Sep-2011 11:03:23] <Hackman238> ion-fan: Yeah, the way I'd suggested wont get you a complete list. What I'll need to do is add a function to iterate the device.os.interfaces and catalog description for each

[28-Sep-2011 11:04:25] <ion-fan> Hackman238: thank you for spending your time on this...

[28-Sep-2011 11:04:47] <Hackman238> ion-fan: Its NP

[28-Sep-2011 11:05:08] <ion-fan> Hackman238: never thought that such a small feature is so much work...

[28-Sep-2011 11:06:07] <Hackman238> ion-fan: Its not too bad. As that pack stands now you can add any attribute, property, zProperty or cProperty to the cataloged search

[28-Sep-2011 11:06:23] <Hackman238> ion-fan: Shouldn't be too bad to add interfaces to that list

[28-Sep-2011 11:08:20] <ion-fan> why do you have to work on the code? i thought i can add device.os.interfaces.interface.description as an attribute?

[28-Sep-2011 11:09:15] <Hackman238> ion-fan: That will only index a single interface unfortunately

[28-Sep-2011 11:09:57] <Hackman238> ion-fan: I'll have to add a lambda function that returns the description list or a method on device that does the same

[28-Sep-2011 11:12:10] <ion-fan> Hackman238: ah i understand: the other attributes are for hosts not for interfaces (which is a level deeper?)

[28-Sep-2011 11:12:25] <Hackman238> ion-fan: thats correct

[28-Sep-2011 11:13:31] <ion-fan> Hackman238: ok happy coding!

[28-Sep-2011 11:13:38] <Hackman238> ion-fan: TY

[28-Sep-2011 11:51:26] <dhopp> hrm...

[28-Sep-2011 11:51:32] <dhopp> Daemon: zenrender not running

[28-Sep-2011 11:51:37] <dhopp> [zenoss:eq2-zen-col-01 ~]$ ps aux | grep zenrender

[28-Sep-2011 11:51:37] <dhopp> zenoss   24440  0.0  0.3 356592 62880 ?        S    10:35   0:00 /opt/zenoss/bin/python /opt/zenoss/Products/ZenRRD/zenrender.py --configfile /opt/zenoss/etc/zenrender.conf --cycle --daemon

[28-Sep-2011 11:51:43] <dhopp> looks like it's running to me

[28-Sep-2011 11:53:21] <Hackman238> dhopp: zenrender stop

[28-Sep-2011 11:53:26] <Hackman238> dhopp: zenrender run now -v10

[28-Sep-2011 11:53:35] <Hackman238> dhopp: Any thing on screen?

[28-Sep-2011 11:55:02] <dhopp> Hackman238:  I got it…that process that was running is a different pid than what was in the pid file (so zenrender stop didn't actually stop it)…I killed that process than started it and it reports ok now

[28-Sep-2011 11:55:31] <Hackman238> Ah that'll do it.

[28-Sep-2011 12:00:01] [disconnected at Wed Sep 28 12:00:01 2011]

[28-Sep-2011 12:00:02] [connected at Wed Sep 28 12:00:02 2011]

[28-Sep-2011 12:00:12] [zenoss-logger (logger bot) has joined #zenoss]

[28-Sep-2011 12:36:02] <dhopp> ok..I created a new Event Class /Status/VIP and assigned a HTTPMonitor data source to use that event class...

[28-Sep-2011 12:36:34] <dhopp> when I run an availability report for that event class it is showing me devices that have HTTPMonitor data sources but those don't have /Status/VIP

[28-Sep-2011 14:14:18] <gloin> I've got a machine throwing bogues "threshold of low swap not met" errors, in 3.0 I can't figure out how to fix the problem.  The machine is barely touching swap but it's extremely minimal so the threshold for it is not appropriate.

[28-Sep-2011 14:14:28] <gloin> s/bogues/bogus

[28-Sep-2011 14:15:03] <Hackman238> gloin: What size is the swap?

[28-Sep-2011 14:16:14] <gloin> Hackman238: one gig

[28-Sep-2011 14:16:27] <gloin> Swap:         1055         71        984

[28-Sep-2011 14:16:42] <gloin> shouldn't be throwing errors

[28-Sep-2011 14:17:17] <gloin> I've tried re-modeling the device but to no avail.

[28-Sep-2011 14:18:25] <Hackman238> gloin: Whats the threshold value?

[28-Sep-2011 14:18:41] <gloin> Hackman238: that's the thing I don't know how to check

[28-Sep-2011 14:19:25] <Hackman238> gloin: Gotcha. Go to the device, pick template from the left

[28-Sep-2011 14:19:45] <Hackman238> gloin: there should be a box on the right, thresholds

[28-Sep-2011 14:19:49] <gloin> Aha

[28-Sep-2011 14:19:57] <Hackman238> gloin: Whats the min value set to?

[28-Sep-2011 14:20:08] <gloin> 102400

[28-Sep-2011 14:20:14] <Hackman238> gloin: thats the problem

[28-Sep-2011 14:20:16] <Hackman238>

[28-Sep-2011 14:20:21] <gloin> indeed so, thanks

[28-Sep-2011 14:20:28] <Hackman238> No problem, anytime

[28-Sep-2011 14:20:49] <gloin> Another question: our storage array uses snapshots, each of which is named SHARENAME_SNAP_SNAPNUMBER

[28-Sep-2011 14:21:01] <Hackman238> gloin: Okay

[28-Sep-2011 14:21:52] <gloin> When the share reaches an alert threshold, each of the snapshots that follows also alerts to that threshold as well, but we don't really care about the snapshots in this context.  Is there any way to filter out *_SNAP_* from the events console or better yet to prevent them from being stored in the first place?

[28-Sep-2011 14:24:01] <gloin> (the main reason we don't care about the snapshots is that they're going to be whatever N% full as the share they were taken from, but all they do is clog the events console with rows of data)

[28-Sep-2011 14:36:33] <Hackman238> gloin: you'll need to write an exvent transform to find the events and evt.drop them

[28-Sep-2011 14:37:45] <Omeganon> I missed the first part of the conversation but if this is using the standard FS monitoring, couldn't he just add '_SNAP_' to the device's configuration properties zFilesSystemMapIgnoreNames?

[28-Sep-2011 14:37:56] <humanix> hi there

[28-Sep-2011 14:37:57] <humanix> hi rmatte

[28-Sep-2011 14:38:05] <Hackman238> Omeganon: Good idea

[28-Sep-2011 14:38:17] <Hackman238> humanix: Hi

[28-Sep-2011 14:38:23] * gloin looks around to figure out how to do that

[28-Sep-2011 14:38:38] <Omeganon> gloin: go to the Device, then Configuration Properties

[28-Sep-2011 14:38:38] <gloin> aha

[28-Sep-2011 14:38:41] <humanix> Hackman238, how are you doing?

[28-Sep-2011 14:38:44] <gloin> Yep, thanks Omeganon!

[28-Sep-2011 14:39:01] <Omeganon> remodel and you should be good to go

[28-Sep-2011 14:39:01] <gloin> So I don't need to use any kind of regex format in that field?

[28-Sep-2011 14:39:10] <Omeganon> not if it's the only value

[28-Sep-2011 14:39:31] <Hackman238> humanix: Busy busy!

[28-Sep-2011 14:39:33] <gloin> It'll be a number of SHARENAME_SNAP_SNAPSHOTNUMBER

[28-Sep-2011 14:39:35] <Hackman238> humanix: yourself?

[28-Sep-2011 14:40:10] <Omeganon> if you want to ignore all of them, just put in _SNAP_ (or other substring common to them all)

[28-Sep-2011 14:40:20] <gloin> Awesome

[28-Sep-2011 14:42:38] <rocket> then remodel

[28-Sep-2011 14:42:40] <gloin> 2011-09-28 11:42:24,141 INFO zen.ZenModeler: Skipping /shares/users_SNAP/2011-09-26.000100 as it matches zFileSystemMapIgnoreNames.

[28-Sep-2011 14:42:45] <gloin> this pleases me greatly

[28-Sep-2011 14:43:42] <gloin> I had to change the ignore but once that was done

[28-Sep-2011 14:43:43] <Hackman238> gloin: Will save on object count and reduce event system load

[28-Sep-2011 14:44:09] <gloin> Hackman238: among other things.  The snapshot full alerts filled the console.  There were about three dozen of them.

[28-Sep-2011 14:44:14] <humanix> jhahaha

[28-Sep-2011 14:44:14] <humanix> same

[28-Sep-2011 14:44:19] <gloin> We know the filesystem is high

[28-Sep-2011 14:44:19] <humanix> Hackman238, haha, same

[28-Sep-2011 14:44:21] <Hackman238> gloin: Yikes. Makes for a messy scene

[28-Sep-2011 14:44:33] <gloin> yeah, kind of masks important events

[28-Sep-2011 14:45:27] <Hackman238> gloin: yep

[28-Sep-2011 14:45:41] <gloin> Is there a way to adjust the alert threshold for a specific filesystem on one device?

[28-Sep-2011 14:46:07] <Hackman238> gloin: You'll need to create a local copy of the template on the device

[28-Sep-2011 14:46:27] <gloin> Yeah, just finding that, should have looked closer before asking.

[28-Sep-2011 14:46:58] <Hackman238> gloin:

[28-Sep-2011 14:49:09] <gloin> Alright, I think I'm figuring this out.  Just for fun, my boss has asked me to compare/contrast Zenoss with Splunk.  I'm thinking that will go beyond apples/oranges though.

[28-Sep-2011 14:49:50] <Hackman238> gloin: Splunk works better for huge volumes of events. Zenoss is certainly more for monitoring.

[28-Sep-2011 14:50:16] <Hackman238> gloin: We use both at Rackspace because they don't negate the others use

[28-Sep-2011 14:55:12] * gloin has kicked ablew in the shins!

[28-Sep-2011 14:56:03] <jmp242> Splunk always seemed expensive to me though

[28-Sep-2011 14:56:14] <gloin> it isn't cheap for sure

[28-Sep-2011 14:56:40] <jmp242> There's so many OSS log correlators so . . .

[28-Sep-2011 14:56:41] <gloin> then again, I was trying to figure out something for one of our web developers and used splunk to get a root cause in all of about thirty seconds

[28-Sep-2011 14:57:08] <gloin> that was this morning

[28-Sep-2011 14:57:21] <gloin> jmp242: what else have you used in that realm that you liked?

[28-Sep-2011 14:57:42] <jmp242> Well, I thought OSSEC is real nice in the version the WUI works with

[28-Sep-2011 14:58:27] <jmp242> I'm still waiting for time and resources to play with logstash some more - elastic search or greylog2 take more server resources than I have right now, but I think would work well vs Splunk - well, for what I would need

[28-Sep-2011 14:58:38] <jmp242> I like OSSEC for the rules so it can do much of the filtering

[28-Sep-2011 14:58:59] <jmp242> but I've never used Splunk, so I don't know if I'm confused about what people use it for

[28-Sep-2011 15:26:48] <dhopp> anybody know why an availability report is showing me devices that have nothing that is in the /Status/VIP (custom event class)?

[28-Sep-2011 15:27:03] <dhopp> I choose that as my event class to report on but I'm seeing other devices

[28-Sep-2011 15:27:31] <Hackman238> dhopp: I dont understand

[28-Sep-2011 15:27:51] <dhopp> Hackman238:  I'm running an availability report (Under Performance Reports)...

[28-Sep-2011 15:28:05] <dhopp> I choose /Status/VIP as the Event class and click 'update'

[28-Sep-2011 15:28:18] <haus1> Splunk has a nice report/graphing structure that slices & dices Zenoss eventy very nicely...

[28-Sep-2011 15:28:37] <haus1> I also use it for OSSEC reporting, they have a plugin for OSSEC

[28-Sep-2011 15:28:42] <dhopp> I thought that would show me only devices/components that actually have something assigned to /Status/VIP

[28-Sep-2011 15:28:53] <Hackman238> dhopp: Thats what I would expect

[28-Sep-2011 15:28:57] <dhopp> haus1:  is that part of the open source version of splunk?

[28-Sep-2011 15:29:06] <dhopp> Hackman238:  that's not what I'm seeing

[28-Sep-2011 15:29:20] <haus1> the "free" under 80mb throtteled version ... yea

[28-Sep-2011 15:29:58] <haus1> you can't point 200 machines at the Free SPLUNK, but ~20 works ...

[28-Sep-2011 15:31:02] <haus1> else the folks at SPLUNK are "VERY PROUD" of their commercial product!!

[28-Sep-2011 15:31:17] <dhopp> Hackman238:  for example I see "EC2Manager" which I know doesn't have /Status/VIP monitoring

[28-Sep-2011 15:31:31] <dhopp> haus1:  yeah I have about 1200 devices

[28-Sep-2011 15:31:56] <Hackman238> dhopp: Im not certain why less its including everything that has an entry in /Status

[28-Sep-2011 15:31:58] <dhopp> and we already have a commercial syslog monitoring package

[28-Sep-2011 15:32:04] <haus1> use OSSEC & Zenoss Core... OSSEC has an OK GUI too

[28-Sep-2011 15:32:29] <dhopp> Hackman238:  hrmph..that would imply that report is broken then

[28-Sep-2011 15:32:37] <haus1> you can do searches from it

[28-Sep-2011 15:32:54] <Hackman238> dhopp: I dont really use that report, I dont know its quirks. Sorry.

[28-Sep-2011 15:33:10] <Hackman238> rocket: Any idea about dhopps problem?

[28-Sep-2011 15:33:25] <Hackman238> haus1: Splunk search is pretty boss

[28-Sep-2011 15:33:28] <dhopp> Hackman238:  I was hoping for something quick and dirty.  Probably in reality I'll have to do a custom report that does availability on a group / device class

[28-Sep-2011 15:33:50] <Hackman238> dhopp: I think theres a pack for it

[28-Sep-2011 15:33:56] <Hackman238> dhopp: Fro group availability

[28-Sep-2011 15:34:00] <haus1> yes, it's powerful, but their licensing is ukky!

[28-Sep-2011 15:34:01] <Hackman238> *for

[28-Sep-2011 15:34:08] <Hackman238> haus1: Aye

[28-Sep-2011 15:34:26] <dhopp> Hackman238:  I know there is one for Resource Utilization

[28-Sep-2011 15:35:28] <haus1> you might be able to do some "event reporting"  with a SNORT instance, but IDK ...

[28-Sep-2011 15:35:47] <haus1> crank up a VM & give it a whirl

[28-Sep-2011 15:36:26] <haus1> that's why I love VirtualBox ... testing simple/easy

[28-Sep-2011 15:36:37] <dhopp> haus1:  really all I need is a report that says the uptime/availability for our VIP checks (since our sla is sort of tied to it)

[28-Sep-2011 15:37:39] <haus1> isn't there a Zenoss report for that ?

[28-Sep-2011 15:38:26] <dhopp> haus1:  there is an availability report but it's not working how I thought

[28-Sep-2011 15:38:40] <haus1> ah... k

[28-Sep-2011 15:38:47] <haus1> I remember seeing it

[28-Sep-2011 15:42:54] <jmp242> haus1: OSSEC WebUI is sadly not compatible with 2.6

[28-Sep-2011 15:43:25] <jmp242> but if it was, I'd just use that myself

[28-Sep-2011 15:44:27] <JohnnyNOC> does anyone else regularly have problems deleting devices?

[28-Sep-2011 15:44:43] <jmp242> I don't reguarily delete devices

[28-Sep-2011 15:44:54] <JohnnyNOC> when i have this problem i normally run this zenfixit.py script that seems to fix my issue..  but i just ran it after having a problem and am trying to delete another device

[28-Sep-2011 15:44:59] <JohnnyNOC> and here i am waiting for zenoss to do something

[28-Sep-2011 15:45:00] <JohnnyNOC> :[

[28-Sep-2011 15:45:56] <Hackman238> JohnnyNOC: Whats the problem?

[28-Sep-2011 15:46:20] <JohnnyNOC> i just regularly have problems deleting devices where zenoss just kind of hangs and the device is never deleted

[28-Sep-2011 15:46:29] <JohnnyNOC> only thing that seems to fix it is running zenfixit.py after trying to delete the device

[28-Sep-2011 15:46:45] <JohnnyNOC> i just ran that, it fixed the device i was trying to delete..  i just tried to delete another and it's having problems again

[28-Sep-2011 15:47:04] <Hackman238> JohnnyNOC: Can you post the trace?

[28-Sep-2011 15:47:04] <JohnnyNOC> well, it complained about problems but the device is gone

[28-Sep-2011 15:47:13] <JohnnyNOC> there is no trace - i deleted via the UI

[28-Sep-2011 15:47:18] <JohnnyNOC> do you delete devices from the command line?

[28-Sep-2011 15:49:48] <Hackman238> JohnnyNOC: I remove them in dmd

[28-Sep-2011 15:50:29] <Hackman238> JohnnyNOC: do d=dmd.Devices.findDevices('devicename') then d.deleteDevice()

[28-Sep-2011 15:51:01] <Hackman238> JohnnyNOC: Then a commit()

[28-Sep-2011 15:51:24] <Hackman238> JohnnyNOC: Safer and less likely to not work without giving you some useful info to fix the prob

[28-Sep-2011 15:51:27] <Hackman238>

[28-Sep-2011 15:56:47] <JohnnyNOC> *nod*

[28-Sep-2011 15:57:00] <JohnnyNOC> guess i'll put together a script for that but damn, it's super annoying when that happens :[

[28-Sep-2011 15:59:27] <Hackman238> JohnnyNOC: Its been cleaner in v4

[28-Sep-2011 16:00:28] <Hackman238> JohnnyNOC: Frankly the problems I have in V4 aren't related to V4 but instead changing our model to fit V4, updating all our custom stuff and cleaning out junk from v1 in our current v2.5.2 instances

[28-Sep-2011 16:00:42] <JohnnyNOC> that's understandable

[28-Sep-2011 16:00:54] <Hackman238> XP is to Windows 7 as v2.5.2 is to v4

[28-Sep-2011 16:01:07] <Hackman238> Both are good, but the new stuff is much better and cleaner

[28-Sep-2011 16:02:53] <dhopp> Hackman238:  I would went with XP is to Linux as 2.5.2 is to v4..but I guess you got your point across :-P

[28-Sep-2011 16:04:08] <JohnnyNOC> i'll wait for v4.0.2

[28-Sep-2011 16:04:10] <JohnnyNOC>

[28-Sep-2011 16:04:15] <Hackman238> dhopp:

[28-Sep-2011 16:04:24] <Hackman238> JohnnyNOC: Wont be v4.0.2 I'm afraid

[28-Sep-2011 16:04:28] <Hackman238> JohnnyNOC: 4.1 or 4.2

[28-Sep-2011 16:05:04] <JohnnyNOC> ok, then v4.1

[28-Sep-2011 16:05:05] <JohnnyNOC> :]

[28-Sep-2011 16:05:06] <Hackman238> JohnnyNOC: Which is in testing for ent now. My guess is shortly after its GA for ent v4.x for core will be in beta

[28-Sep-2011 16:05:34] <JohnnyNOC> cool

[28-Sep-2011 16:05:43] <JohnnyNOC> we've only been on 3.1.0 for a couple 1-2 months now

[28-Sep-2011 16:05:46] <JohnnyNOC> no rush

[28-Sep-2011 16:06:17] <davetoo> 4.0.2 exists, but not sure what "domain" you're talking about.

[28-Sep-2011 16:06:59] <Hackman238> davetoo: 4.0.2 ent, not core.

[28-Sep-2011 16:07:19] <davetoo> oh, are we going to release a 4.x core?

[28-Sep-2011 16:07:21] <Hackman238> davetoo: est core (far as I have heard) is 4.1 or 4.2

[28-Sep-2011 16:07:38] <Hackman238> davetoo: Yes. Its on the ZCA road map too

[28-Sep-2011 16:07:46] <Hackman238> ...need to get that map better exposure

[28-Sep-2011 16:07:49] <Hackman238>

[28-Sep-2011 16:07:50] <davetoo> I haven't had time to keep up with that question.

[28-Sep-2011 16:08:15] <davetoo> I've been too busy learning about the *other* ZCA

[28-Sep-2011 16:09:03] <Hackman238> davetoo:

[28-Sep-2011 16:09:34] <davetoo> i.e. figuring out how to use interfaces/components/zcml

[28-Sep-2011 16:10:21] <Hackman238> davetoo: Yes, important areas.

[28-Sep-2011 16:10:42] <Hackman238> davetoo: If you like I have a zope programming PDF that has some great ZCA info

[28-Sep-2011 16:10:59] <Hackman238> davetoo: You work for Zenoss?...I remember you, but at the same time I dont/

[28-Sep-2011 16:11:03] <davetoo> I probably have it

[28-Sep-2011 16:11:06] <davetoo> Muthakaden?

[28-Sep-2011 16:11:13] <Hackman238> davetoo: Oh gotcha

[28-Sep-2011 16:11:17] <Hackman238> davetoo: sorry

[28-Sep-2011 16:12:19] <davetoo> just meant that I've done enough digging that unless I've missed something... you are talking about that muthakaden book?

[28-Sep-2011 16:13:01] <davetoo> I also have the Zope 3 dev handbook ( tree and electron versions )

[28-Sep-2011 16:13:06] <Hackman238> davetoo: I am

[28-Sep-2011 16:15:03] <dhopp> so on this system /opt is a symlink to /usr/local..if I run zendmd I get the following warnings:

[28-Sep-2011 16:15:04] <dhopp> 2011-09-28 16:14:36 WARNING Application Duplicate Product name: After loading Product 'DataCollector' from '/opt/zenoss/Products', I skipped the one in '/usr/local/zenoss/Products'.

[28-Sep-2011 16:15:19] <dhopp> which I know is because of the symlink..any way to suppress those?  It's not a big deal, just kind of annoying

[28-Sep-2011 16:15:45] <davetoo> change the symlink to a bind/loopback mount?

[28-Sep-2011 16:16:09] <dhopp> davetoo:  hrm..

[28-Sep-2011 16:16:13] <dhopp> might try that

[28-Sep-2011 16:16:16] <davetoo> You won't get rid of those warnings without a lot of work

[28-Sep-2011 16:16:37] <dhopp> davetoo:  like I said..not a big deal but kind of annoying

[28-Sep-2011 16:16:52] <davetoo> well, what would worry me is the root cause;

[28-Sep-2011 16:17:00] <davetoo> something is ignoring the symlink.

[28-Sep-2011 16:19:04] <davetoo> I recenly used bind-mounts like this on some new production systems; it's working fine.  (the bind mount was actually a new concept to me until a couple of months ago; I don't keep up with the kernel/core utils as much as I should.)

[28-Sep-2011 16:19:22] <dhopp> in my limited research this seems to be common with python/zope when using symlinks..

[28-Sep-2011 16:19:45] <dhopp> I'm familiar with bind mounts (we use them when we chroot jail a user but want them to have access to something outside of the jail)

[28-Sep-2011 16:19:46] <davetoo> zope.  Wouldn't be inherent to python,.

[28-Sep-2011 16:21:03] <davetoo> In these cases, what we had was a SAN volume attached to a VM.

[28-Sep-2011 16:21:53] <davetoo> well two; one with the canonical mount set to noatime, and the second, noatime+data=writeback.

[28-Sep-2011 16:22:48] <dhopp> I could probably break the symlink and then move data from /usr/local/ to /opt that needs to be there but not sure if it's worth the effort

[28-Sep-2011 16:26:46] <Hackman238> Correction all, Core will be v4.2

[28-Sep-2011 16:34:09] <jvt5009> Anyone know why I can't edit a manually added component interface?  I go to details and see the information but no textboxes and no way of editing. Save/Cancel are grayed out.

[28-Sep-2011 16:34:37] <jvt5009> I meant an IP Interface

[28-Sep-2011 16:43:25] <rocket> jvt5009: which version?

[28-Sep-2011 16:43:58] <jvt5009> 3.2.0

[28-Sep-2011 16:52:37] <Hackman238> rocket: Turns out our cloud folks are dumping solarwinds for Zenoss

[28-Sep-2011 16:53:00] <Hackman238> rocket: ...so what we'll be monitoring is about to multiply by 15 or 20

[28-Sep-2011 16:53:42] <jvt5009> nm think I found my fix http://dev.zenoss.com/trac/ticket/7869

[28-Sep-2011 16:54:25] <rocket> jvt5009: thats the one .. I was just on a customer call before I could tell you

[28-Sep-2011 16:54:32] <rocket> Hackman238: wheeee

[28-Sep-2011 16:54:46] <jvt5009> ty

[28-Sep-2011 16:56:04] <dhopp> Hackman238:  and you are going to do that with core right? ;-)

[28-Sep-2011 16:57:07] <Hackman238> dhopp: Its undecided. Zenoss will have to throw cheap enough pricing at us

[28-Sep-2011 16:58:22] <rocket> now thats just teasing

[28-Sep-2011 16:59:38] <Hackman238> rocket: LOL Its not really up to me. Technically unofficial, but likely.

[28-Sep-2011 17:00:02] <Hackman238> Whoa.

[28-Sep-2011 17:00:05] <Hackman238> rocket: Chekc this out

[28-Sep-2011 17:00:08] <Hackman238> ERROR:zen.UUID:Encountered a guid exception

[28-Sep-2011 17:00:09] <Hackman238> Traceback (most recent call last):

[28-Sep-2011 17:00:09] <Hackman238>   File "/opt/zenoss/Products/ZenUtils/guid/event.py", line 42, in registerGUIDToPathMapping

[28-Sep-2011 17:00:12] <Hackman238>     catalog = object.global_catalog

[28-Sep-2011 17:00:14] <Hackman238> AttributeError: global_catalog

[28-Sep-2011 17:00:36] <rocket> where was that?  thats new

[28-Sep-2011 17:00:43] <rocket> a migrate step?

[28-Sep-2011 17:00:58] <Hackman238> rocket: Yeah, from london Dc.

[28-Sep-2011 17:01:06] <Hackman238> rocket: I'm not even sure what it means

[28-Sep-2011 17:01:29] <rocket> can it wait till the morning I am about to step out or is it critical?

[28-Sep-2011 17:01:34] <rmatte> aloha people

[28-Sep-2011 17:01:39] <Hackman238> rocket: It can wait

[28-Sep-2011 17:01:47] <Hackman238> rmatte: Hey, how goes it?

[28-Sep-2011 17:02:02] <rocket> ok open a ticket so we can track it .. but lets get it in the morning I guess

[28-Sep-2011 17:02:13] <rocket> something is funky with your global catalog I think

[28-Sep-2011 17:02:24] <Hackman238> rocket: Alrighty

[28-Sep-2011 17:02:26] <rocket> you might want to rebuild the global catalog

[28-Sep-2011 17:02:32] <rocket> by rerunning zencatalog

[28-Sep-2011 17:02:53] <Hackman238> rocket: I tried already

[28-Sep-2011 17:03:00] <rmatte> Hackman238: good, worked on a change from 8:30pm last night until 7am this morning after a full day of work.  Took the day off.

[28-Sep-2011 17:03:08] <dhopp> ok..any cisco pix gurus here?

[28-Sep-2011 17:03:15] <rmatte> ...at a bar downtown right now on my laptop having some brews

[28-Sep-2011 17:03:15] <Hackman238> rmatte: Wow.

[28-Sep-2011 17:03:24] <rocket> Hackman238: well we will get it in the morning

[28-Sep-2011 17:03:26] <Hackman238> rmatte: And you're here in the Zenoss forum?

[28-Sep-2011 17:03:32] <Hackman238> rocket: Alrighty. NP

[28-Sep-2011 17:03:38] <Hackman238> rmatte:

[28-Sep-2011 17:03:43] <rmatte> yup, figured people would be around

[28-Sep-2011 17:03:50] <rmatte> and technically this is the channel

[28-Sep-2011 17:04:37] <Hackman238> rmatte: That it is! I knew #zenoss was 'cool'

[28-Sep-2011 17:04:45] <rmatte> and yes I am, the benefits of haing a laptop

[28-Sep-2011 17:04:56] <rmatte> lol

[28-Sep-2011 17:05:12] <rmatte> smaller laptop though so the keyboard is a bit cramped

[28-Sep-2011 17:05:15] <Hackman238> that said, its 16:00h, I have to bounce

[28-Sep-2011 17:05:20] <Hackman238> rmatte: LOL

[28-Sep-2011 17:05:20] <rmatte> later

[28-Sep-2011 17:05:23] <Hackman238> Later all!

[28-Sep-2011 17:05:49] <dhopp> rmatte:  haing?  how many brews have you had?

[28-Sep-2011 17:05:52] <dhopp> :-P

[28-Sep-2011 17:06:02] <rmatte> I'm on my second honestly

[28-Sep-2011 17:06:33] <rmatte> deal on mini pitchers of my favourite beer today though

[28-Sep-2011 17:07:19] <dhopp> rmatte:  so calling into work tomorrow?

[28-Sep-2011 17:08:19] <rmatte> nah, won't be out that late lol

[28-Sep-2011 17:08:36] <rmatte> my sleep schedule is crazy messed from that change though

[28-Sep-2011 17:08:44] <rmatte> slept from 7am to 3pm

[28-Sep-2011 17:08:53] <rmatte> and it's 5pm right now

[28-Sep-2011 17:08:58] <rmatte> beer for bfast,yum

[28-Sep-2011 17:09:07] <dhopp> breakfast of champions!

[28-Sep-2011 17:09:14] <rmatte> indeed

[28-Sep-2011 17:09:42] <dhopp> breakfast is the most important meal of the day so be sure not to drink the lite beer stuff…you need to get your calories

[28-Sep-2011 17:10:19] <rmatte> having some india pale ale at the moment

[28-Sep-2011 17:10:41] <rmatte> Alexander Keith's - "The pride of Nova Scotia"

[28-Sep-2011 17:11:06] <rmatte> also one of the top selling beers in Canada

[28-Sep-2011 17:11:34] <jvt5009> will adding an IPInterface component to an existing device prevent that IP from auto detecting as another device?

[28-Sep-2011 17:12:23] <rmatte> jvt5009: yes, unfortunately.  You can manually add another device with another IP and then reset the IP to that one though to work around it

[28-Sep-2011 17:12:37] <rmatte> like add it with a fake IP then reset the IP

[28-Sep-2011 17:13:00] <jvt5009> actually I don't want it to appear as a seperate device.  Thanks

[28-Sep-2011 17:13:08] <rmatte> ah, no problem

[28-Sep-2011 17:14:10] <rmatte> (I should add that I've never tried it with manually addinga component, I've only seen the behaviour with actual modeled interfaces

[28-Sep-2011 17:14:15] <rmatte> but in theory it should

[28-Sep-2011 17:14:39] <jvt5009> k, I'll keep an eye on it just in case

[28-Sep-2011 19:54:56] kocolosk|away is now known as kocolosk

[28-Sep-2011 21:36:59] kocolosk is now known as kocolosk|away

[28-Sep-2011 21:37:26] kocolosk|away is now known as kocolosk

[28-Sep-2011 21:43:10] kocolosk is now known as kocolosk|away

[28-Sep-2011 22:18:38] kocolosk|away is now known as kocolosk

[29-Sep-2011 00:00:01] [disconnected at Thu Sep 29 00:00:01 2011]

[29-Sep-2011 00:00:01] [connected at Thu Sep 29 00:00:01 2011]

[29-Sep-2011 00:00:11] [zenoss-logger (logger bot) has joined #zenoss]

[29-Sep-2011 00:35:32] Omeganon_ is now known as Omeganon

[29-Sep-2011 02:26:11] kocolosk is now known as kocolosk|away

[29-Sep-2011 08:00:41] <jmp242> Morning all

[29-Sep-2011 08:11:51] <ion-fan> good morning

[29-Sep-2011 08:40:16] <Hackman238> Morning all!

[29-Sep-2011 08:53:27] <Hackman238> This --> http://m3report.files.wordpress.com/2011/07/atf-cartoon.jpg <--

[29-Sep-2011 09:13:33] <jb> hmm

[29-Sep-2011 09:13:35] <jb> morning

[29-Sep-2011 09:15:22] <Hackman238> jb: Hey, how goes it?

[29-Sep-2011 09:15:27] <jb> not bad.

[29-Sep-2011 09:15:29] <jb> got a question..

[29-Sep-2011 09:15:33] <jb> http://pastebin.com/x2AsMF86

[29-Sep-2011 09:15:53] <jb> what do you think the best way to parse those stats (outgoing queries by type, etc) would be and throw them at zenoss as a datasource?

[29-Sep-2011 09:16:00] <jb> these reside in a file on the client machine

[29-Sep-2011 09:16:11] <Hackman238> jb: I have a guy you can contact that does this now. I'm sure he would share his solution

[29-Sep-2011 09:16:15] <jb> i typically used net-snmp's exec and a script on the client machine to push them into snmp.

[29-Sep-2011 09:16:22] <jb> Hackman238: already talked to him

[29-Sep-2011 09:16:25] <jb> we have the same solution

[29-Sep-2011 09:16:32] <jb> this is for a newer bind though, where there are many more stats

[29-Sep-2011 09:16:36] <Hackman238> *face palm* sorry

[29-Sep-2011 09:16:39] <jb> np

[29-Sep-2011 09:16:45] <Hackman238> jb: Very short memory

[29-Sep-2011 09:17:03] <jb> i think a client side script that parsed that stat output and pushed each and every number thru snmp would be too much.

[29-Sep-2011 09:17:19] <jb> since there will be 20+ datasources

[29-Sep-2011 09:17:20] <Hackman238> jb: Personally what I would do is write a custom parser plugin for zenoss command datasources or snmp datasources then fetch the data via that method

[29-Sep-2011 09:18:08] <jb> how would you fetch though?

[29-Sep-2011 09:18:12] <rocket> ssh

[29-Sep-2011 09:18:28] <rocket> use zencommand ... cat /path/to/foo

[29-Sep-2011 09:18:35] <rocket> use the plugin to parse the result set

[29-Sep-2011 09:18:41] <jb> sup rocket

[29-Sep-2011 09:18:45] <rocket> and save the datapoints

[29-Sep-2011 09:18:55] <jb> ok so make zencommand use ssh.

[29-Sep-2011 09:18:58] <ion-fan> hackman238: how is the status with the device search zenpack?

[29-Sep-2011 09:18:59] <rocket> hey jb

[29-Sep-2011 09:19:33] <jb> rocket: can zencommand use ssh natively?  or would I need zencommand to call a script that lives on the zenoss box to do the ssh work?

[29-Sep-2011 09:20:05] <rocket> it can do it natively

[29-Sep-2011 09:21:41] <jb> cool, BIND just appends each stat routine to the file, it doesn't overwrite, so I'd also need to check for start/end lines.

[29-Sep-2011 09:23:01] <rocket> well just grab the whole file

[29-Sep-2011 09:23:19] <rocket> and then have the parser handle that

[29-Sep-2011 09:24:38] <jb> ok, so I would need to write a custom parser, correct?

[29-Sep-2011 09:26:46] <rocket> yes

[29-Sep-2011 09:27:05] <rocket> or you can do a nagios style script in bash

[29-Sep-2011 09:27:32] <rocket> but that would require putting the script on the remote server being monitored

[29-Sep-2011 09:27:33] <dhopp> jb: are you trying to get stats out of bind?

[29-Sep-2011 09:27:33] <jb> thats probably more of my level

[29-Sep-2011 09:27:38] <jb> dhopp: yes.

[29-Sep-2011 09:27:47] <jb> dhopp: it was easy with < bind 9.4.

[29-Sep-2011 09:28:02] <jb> but 9.5+ offers a ton of additional datasources/points.

[29-Sep-2011 09:28:21] <dhopp> jb: http://www.jasonmcneil.net/blog/extend-snmp-and-your-zenoss-monitoring-capabilities/

[29-Sep-2011 09:28:41] <jb> hrm

[29-Sep-2011 09:28:43] <jb> who is that guy?

[29-Sep-2011 09:29:08] <jb> cause I wrote a pretty  much identical blog post

[29-Sep-2011 09:30:01] <dhopp> jb: haha

[29-Sep-2011 09:30:14] <dhopp> jb: I just found that article yesterday when I was wondering how to do the same thing

[29-Sep-2011 09:30:36] <jb> rocket: so if the nagios plugin returned something like "10 36 3999 1990," I could tell zen that $1 was 'A' queries, $2 is 'MX' queriest and so on?

[29-Sep-2011 09:30:49] <rocket> jb: if you can figure out the parser I would recommend doing that .. that way you do not need to install anything on the remote side

[29-Sep-2011 09:31:03] <rocket> you would need it to be the format

[29-Sep-2011 09:31:23] <Hackman238> Sorry, all. Unexpectadly I have to run. I'll be back.

[29-Sep-2011 09:31:33] <rocket> OK|a=10 mx=36 foo=3999 bar=1990

[29-Sep-2011 09:31:44] <rocket> then your datapoints will be a,mx,foo,bar

[29-Sep-2011 09:32:01] <rocket> parser will be auto or nagios

[29-Sep-2011 09:32:27] <rocket> click the use ssh check box and then call the path to the nagios script on the remote machine

[29-Sep-2011 09:34:34] <jb> ahh pefect

[29-Sep-2011 09:34:41] <jb> i may give that s hot.

[29-Sep-2011 09:34:43] <jb> thanks

[29-Sep-2011 10:47:47] <nyeates> anyone home?

[29-Sep-2011 10:47:55] <jrh0090> Hey Nick

[29-Sep-2011 10:48:08] <nyeates> Hello there

[29-Sep-2011 10:56:06] <dhopp> knock knock

[29-Sep-2011 10:56:35] <nyeates> Hola dhopp

[29-Sep-2011 10:56:38] <nyeates> what can we do you for

[29-Sep-2011 10:58:26] <dhopp> nyeates:  do you know if it's possible to have ldap users and groups but then assign a higher role locally?  For example I have an ldap group that will be assigned ZenManger role, but would like to keep Manager for very specific users

[29-Sep-2011 10:58:34] jgartman_ is now known as jgartman

[29-Sep-2011 10:59:00] <nyeates> this is in core or ent?

[29-Sep-2011 10:59:19] <dhopp> I know I could create another LDAP group, but there are people that have Domain Admin access that I might not want to have Manager access and if they have Domain Admin they could just put themselves in that group...

[29-Sep-2011 10:59:19] <dhopp> core

[29-Sep-2011 10:59:29] <jrh0090> I think if you chanage the role in zenoss it won't be propagated to the ldap group, it is local to zenoss only

[29-Sep-2011 11:00:08] <dhopp> I can get an ldap user to have a local role..but that's if I'm not using ldap groups...

[29-Sep-2011 11:00:23] <dhopp> it's not a huge deal if it's not possible..but just wondering if it was

[29-Sep-2011 11:01:09] <nyeates> im not as clear on core, but i think its similar....like jrh said, if there is a local zenoss user, just set them as Manager and it will override....so it may be a manual process on the zenoss side

[29-Sep-2011 11:01:23] <nyeates> im not sure this answers what u are asking?

[29-Sep-2011 11:02:26] <dhopp> nyeates:  basically what I'm wondering is if I have a ZenManager ldap group when a user logs in they get the ZenManager role..which works..but then I would like the Manager role to be controlled locally

[29-Sep-2011 11:02:59] <nyeates> dhopp: do you have issue with using LDAP groups AND setting a local zenoss role on someone?

[29-Sep-2011 11:03:07] <jrh0090> have you tried changing it in zenoss and seeing what happens?

[29-Sep-2011 11:03:11] <nyeates> dhopp: or you are wondering how to go about doing this?

[29-Sep-2011 11:03:48] <dhopp> nyeates:  it doesn't seem to be working…but I think I have something screwed up because now it does't seem my ldap groups are working at all…need to go check the configuration

[29-Sep-2011 11:06:12] <nyeates> dhopp: make sure that both ldap and local auth plugins are enabled on the backend zope /manage interface....not one or the other...i think u want ldap first in the list and local auth second...send a screenshot of auth plugin setup if you take a look

[29-Sep-2011 11:07:13] <dhopp> I have both enabled…I can login as an ldap user..but it doesn't seem to be setting my role based on my group membership.  I had this working before and was playing around with the configuration so I'm sure I screwed something up

[29-Sep-2011 11:07:48] <orudie> is zenoss used for load balancing ?

[29-Sep-2011 11:08:00] <dhopp> orudie:  what do you mean?

[29-Sep-2011 11:08:02] <davetoo> "used for load balancing?

[29-Sep-2011 11:09:25] <dhopp> ha!  ldap group mappings will only work if you assign the users to the right groups :-P

[29-Sep-2011 11:09:27] <orudie> well when someone asks you a question "have you ever worked with load balancing, and what tools did you use?" what would be the apropriate answer ?

[29-Sep-2011 11:09:28] <davetoo> If you mean "can it be managed by Heartbeat/Pacemaker", yes, though I haven't actually seen a KB article/how-to that describes doing it with Pacemaker.  You'd have to write a pacemaker resource agent; the init.d is not quite perfectly LSB complient.

[29-Sep-2011 11:09:50] <davetoo> Zenoss itself has nothing to do with load-balancing.

[29-Sep-2011 11:10:12] <davetoo> If you mean as part of a loadsharing/dispatching system, ... way too heavyweight/coarse-grained.

[29-Sep-2011 11:12:02] <dhopp> orudie:  zenoss is a monitoring packaging.  It could monitor a load balanced system. you can use distributed collectors to spread the zenoss load but that isn't load balancing

[29-Sep-2011 11:13:25] <nyeates> orudie: maybe you heard that someone was using zenoss, as one tool in many, to complete their entire load balance solution. Our part in that tool-chain would be monitoring the load balanced system to assure that it is operating as it should be.

[29-Sep-2011 11:13:48] <Jane_Curry> hi guys...

[29-Sep-2011 11:13:57] <orudie> nyeates, thanks

[29-Sep-2011 11:14:08] <Hackman238> How to address this case: 'device name' exists in '/Discovered'; zendmd: .moveDevice(d,'/New Class'); Device d ('Device name') now has a ton of broken releationships.; zenchkrel and .checkRelations(repair=True) finds broken renationships and acts, but fails to fix all the relationships.; d.getPrimaryParent != d.getPrimaryPath and d.getPrimaryParent results in primaryAq key error.; Device instance recreation does not fix problem; d.deleteDe

[29-Sep-2011 11:14:30] <nyeates> Hi Jane

[29-Sep-2011 11:14:47] <davetoo> Hackman: pastebin?

[29-Sep-2011 11:14:47] <Jane_Curry> nasty - was this the guy you were helping out yesterday????

[29-Sep-2011 11:15:02] <Hackman238> Jane_Curry: Aye

[29-Sep-2011 11:15:27] <Jane_Curry> Just for a laugh, Are there any GALS out there among the guys??????????

[29-Sep-2011 11:15:47] <themactech> No, but i can cook, is that worth something?

[29-Sep-2011 11:15:54] <nyeates> lol

[29-Sep-2011 11:16:03] <Jane_Curry> Lots of brownie points for that themactech

[29-Sep-2011 11:16:07] <jrh0090> so the guy tried doing ._delOb on the broken relationships ?

[29-Sep-2011 11:16:14] <davetoo> Hackman238: is that your system?  Or a client's?

[29-Sep-2011 11:16:38] <Hackman238> davetoo: >>> dev.deleteDevice() Traceback (most recent call last):   File "<console>", line 1, in  <module>   File "/opt/zenoss/Products/ZenModel/Device.py", line 1842, in deleteDevice      parent = self.getPrimaryParent()   File  "/opt/zenoss/Products/ZenRelations/PrimaryPathObjectManager.py", line 85, in getPrimaryParent

[29-Sep-2011 11:16:38] Spawn is now known as Guest97443

[29-Sep-2011 11:16:40] <nyeates> Jane_Curry: not a lot of people on today, so bad day to ask

[29-Sep-2011 11:16:43] <Hackman238>  return self.__primaryparent__.primaryAq() AttributeError: 'NoneType' object has no atptribute  'primaryAq'

[29-Sep-2011 11:16:53] <Hackman238> davetoo: a community member

[29-Sep-2011 11:16:57] <jrh0090> wow that is bad

[29-Sep-2011 11:17:09] <Hackman238> jrh0090: Aye

[29-Sep-2011 11:17:23] <davetoo> whatever it is, it's an orphan

[29-Sep-2011 11:17:29] <dhopp> nyeates:  it works now that I put the users in the right groups :-) and local roles override the ldap mapped roles

[29-Sep-2011 11:17:40] <Hackman238> jrh0090: Tried to recreate instance via dmd and commit. Result was then http://pastebin.com/QQcyZSYL

[29-Sep-2011 11:18:06] <nyeates> dhopp: "now that i put the userse in the right groups".... ldap groups?

[29-Sep-2011 11:18:45] <davetoo> Hackman238: my guess is that somehow the device belonged to a Device Class/organizer that was deleted, but something kept the device itself from being cleaned up.

[29-Sep-2011 11:18:54] <Hackman238> jrh0090/davetoo: used this meathod to recerate:

[29-Sep-2011 11:18:55] <davetoo> what version of zenoss?

[29-Sep-2011 11:18:56] <Hackman238> dVc = dmd.getDmdRoot("Devices").createOrganizer('/Discovered')

[29-Sep-2011 11:18:56] <Hackman238> dVn = 'dV name'

[29-Sep-2011 11:18:56] <Hackman238> dV = dVc.createInstance(dVn)

[29-Sep-2011 11:18:56] <Hackman238> dV.manage_editDevice('', '', '', '161', '', '', -1, '', '', '', '', '', '', '', '', 'localhost', 0)

[29-Sep-2011 11:18:58] <dhopp> nyeates:  yeah…I put my users in groups that ZMI wasn't configured to look at…so the mappings weren't working

[29-Sep-2011 11:18:59] <dhopp> heh

[29-Sep-2011 11:18:59] <Hackman238> commit()

[29-Sep-2011 11:19:09] <Hackman238> davetoo: v3.1

[29-Sep-2011 11:19:22] <jrh0090> Maybe try: md.Devices.Discovered._delOb(dV.id) instead, if I remember correctly that actually deletes without firing events

[29-Sep-2011 11:19:32] <davetoo> Why createOrganizer?

[29-Sep-2011 11:19:33] <themactech> Jane, I am still working on a custom component zenpack, and using your document to try and figure it out.  I have a question.  On page 44 when I write the equivalent of the BridgeInterface.py for my component, in the "custom variables from modeling"  do I only put data I will get from modeling or also the data point for the monitoring?

[29-Sep-2011 11:19:38] <davetoo> did they delete /Discovered?

[29-Sep-2011 11:19:46] <nyeates> dhopp: ahh, that will do it.... yeah ussually if you fiddle with the settings long enough, you get stuff to work :-)

[29-Sep-2011 11:19:48] <jrh0090> or if it is a .devices. relation ship you can manually remove it from dmd.Devices.Discovered.devices._objects

[29-Sep-2011 11:19:49] <Hackman238> davetoo: negative, but if exists it assigns

[29-Sep-2011 11:20:01] <rocket> guys be very careful of the _delOb methods it only works on half of the relationship etc

[29-Sep-2011 11:20:09] <rocket> Hackman238: it sounds like a support ticket to me

[29-Sep-2011 11:20:13] <jrh0090> that is true

[29-Sep-2011 11:20:20] <dhopp> nyeates:  but I have another question…I created /Status/VIP event class and assigned a monitoring template datasource to use that event class.  When I run the availability report and choose that event class I see devices that are not using the new template…if I choose /Status/VIP shouldn't the availability only show me stuff that is using that event class?

[29-Sep-2011 11:20:24] <Hackman238> rocket: Good to know, i didnt know that

[29-Sep-2011 11:20:42] <davetoo> yeah depending on what kind of relation it is, you may have to go at it from both ends.

[29-Sep-2011 11:20:51] <rocket> yeah it does not automatically cleanup the other half of the relationship

[29-Sep-2011 11:21:38] <Hackman238> davetoo: I tried to .removeRelation on the object to cut it off and let it float int he void, but then this: http://pastebin.com/4TuZqTRv

[29-Sep-2011 11:21:41] <jrh0090> dhopp have new events come in for devices with that event class? It might just be querying the current state of the status table

[29-Sep-2011 11:21:43] <dhopp> davetoo:  how do you know what kind of relationship requires that?

[29-Sep-2011 11:21:53] <dhopp> jrh0090:  yes

[29-Sep-2011 11:22:43] <jrh0090> hmm, i don't know it should hve picked up those devices

[29-Sep-2011 11:22:51] <rocket> dhopp: when you look at the dev guides its the ones with bidirectional links

[29-Sep-2011 11:23:04] <dhopp> jrh0090:  it sees the devices that has that class..but it also sees other devices (like EC2Manager)

[29-Sep-2011 11:23:19] <dhopp> rocket:  gotcha

[29-Sep-2011 11:23:27] <davetoo> probably the ...Cont ones, eh?  I still don't have those all in my head.

[29-Sep-2011 11:23:33] <rocket> zenchkrels is actually testing the conditions if it finds half a link from a container object it will re-add it etc

[29-Sep-2011 11:23:33] <davetoo> the containers.

[29-Sep-2011 11:24:00] <rocket> if there is a half of a link from an object that is not a container it will remove that relationship etc ..

[29-Sep-2011 11:24:10] <rocket> but there are cases when it cant figure things out

[29-Sep-2011 11:24:12] <Hackman238> rocket / davetoo  / jrh0090: was my use of removeReltation valid?

[29-Sep-2011 11:24:35] <Hackman238> I figured remove all the rels, try to .checkRelations(repair=True) it, if fail, cut again and let it float.

[29-Sep-2011 11:24:47] <rocket> it should have been I believe ..  I dont typically use remove relation myself

[29-Sep-2011 11:24:53] <nyeates> dhopp: maybe make a nice clear replicatable trac bug case on this....put it at like priority 4 and say it needs to be verified

[29-Sep-2011 11:25:00] <jrh0090> as far as i know it should be

[29-Sep-2011 11:25:14] <dhopp> nyeates:  not the answer I was hoping for :-P

[29-Sep-2011 11:25:16] <rocket> Hackman238: did you use my device repair script I sent you?

[29-Sep-2011 11:25:37] * davetoo wonders if even then, a catalog index might keep a ref to the object and keep it from being garbage-collected, and maybe cause problems?

[29-Sep-2011 11:25:39] <Hackman238> rocket: Yes, let me find that pastebin

[29-Sep-2011 11:25:50] <nyeates> dhopp: i dont really know enough about that report to give you anymore....i think it should do what you are thinking...so sounds like bug to me

[29-Sep-2011 11:26:37] <dhopp> nyeates:  actually for that matter it seems it doesn't matter what Event Class I choose it always shows the same devices

[29-Sep-2011 11:26:50] <jrh0090> yeah sounds like a bug to me too, I am looking at the availability code and it should be pulling those devices if there are events for that device with that event class or a child of that event class

[29-Sep-2011 11:26:58] <Hackman238> rocket: it died silently when reaching one of the problem devices (10.102.232.255) http://pastebin.com/ZDYZ3UCf

[29-Sep-2011 11:27:19] <jb> rocket: erm, can I use key based SSH auth with zencommand?

[29-Sep-2011 11:27:22] <jrh0090> maybe that filter is not making it through, are the other filters working ?

[29-Sep-2011 11:27:39] <Hackman238> davetoo: I did uncatalog it first before all this, though IDK if it became recataloged

[29-Sep-2011 11:28:07] <rocket> jb: yes

[29-Sep-2011 11:28:11] <Hackman238> Sorry to bring such a terrible question!

[29-Sep-2011 11:28:32] <jrh0090> hah, it's ok. That has to be one of the worst ZODB states I have seen

[29-Sep-2011 11:29:00] <Hackman238> jrh0090: yeah the user was using moveObject to change classes instead of changeDeviceClass

[29-Sep-2011 11:29:04] <jb> rocket: its the zenoss user that executes the SSH, correct?

[29-Sep-2011 11:29:34] <dhopp> jb: it will attempt to login with the zCommandUser I believe

[29-Sep-2011 11:29:58] <jb> just trying to figure out who's public key on the zenoss machine I need to deploy to my remote hosts

[29-Sep-2011 11:30:01] <rocket> jb it is .. its not using a ssh client its using a python ssh library

[29-Sep-2011 11:30:04] <jrh0090> I don't know if it is helpful but there is a device facade method to move to help out now: something like facade = getFacade('device') facade.moveDevices(uids, organizerUid)

[29-Sep-2011 11:30:09] <jb> k

[29-Sep-2011 11:30:22] <rocket> jb this should be noted in the documentation

[29-Sep-2011 11:30:24] <jrh0090> hopefully that will be safe and not leave the ZODB in a messed up state

[29-Sep-2011 11:30:25] <jb> sorry

[29-Sep-2011 11:30:40] <Hackman238> jrh0090: Nice

[29-Sep-2011 11:31:17] <Hackman238> jrh0090: I know changeDeviceClass() is generally safe, I think moveObject jsut hammed up all the rels

[29-Sep-2011 11:31:34] <jrh0090> sounds like it

[29-Sep-2011 11:32:06] <Hackman238> Alrighty, I just wanted to be sure there wasn't a better solution that we came up with

[29-Sep-2011 11:32:20] <Hackman238> Mostly comprising of use the correct method. LOL

[29-Sep-2011 11:33:06] <jrh0090> hah yeah hopefully he can get that relationship cleared up, I would hate for him to have to do a reinstall because of that

[29-Sep-2011 11:33:49] <rocket> Hackman238: I believe my device repair script will fix this device

[29-Sep-2011 11:34:18] <Hackman238> jrh0090: Fortuantely the rollout was minimal enoguh that a devicedump and reinstall wasnt a big deal

[29-Sep-2011 11:34:25] <dhopp> rocket:  Device repair script?  can you share?

[29-Sep-2011 11:34:33] <nyeates> Community ZenPack devs: I am working hard to convert all ~200 community zenpacks over to individual seperate github repos. Going to be accompanied by improved documentation around how to submit ZPs on github. I plan to take phonegi's awesome doc and make it into an updated html version.

[29-Sep-2011 11:34:34] <Hackman238> rocket: it died silently when reaching one of the problem devices (10.102.232.255)  http://pastebin.com/ZDYZ3UCf

[29-Sep-2011 11:34:36] <davetoo> combined with a more agressive backup cron job?

[29-Sep-2011 11:35:18] <rocket> dhopp: sure .. it fixes many situations re broken devices

[29-Sep-2011 11:35:24] <nyeates> Community ZenPack devs: github is moving to be our central way of handling ZPs....we plan to automate stuff and I imagine that it will surround around git and github. If you dont know it yet, brush up on your git skills.

[29-Sep-2011 11:36:28] <dhopp> nyeates:  git?  man my company is just now trying to move from cvs to svn..:-P

[29-Sep-2011 11:37:03] <nyeates> dhopp: I wont comment there haha

[29-Sep-2011 11:37:27] <Jane_Curry> nyeates: sounds good - what about a training session on git - can we get phonegi to do it???????

[29-Sep-2011 11:39:16] <rocket> dhopp: http://pastebin.com/NTtrpAQS

[29-Sep-2011 11:39:18] <nyeates> Jane_Curry: a step at a time....lets get it all out there first and then we should talk about this.. If training will commense, it will likely be me running it. I think it sounds like a good idea. Maybe at the least I could plan to do a live/recorded step through of the instructions that i come up with

[29-Sep-2011 11:40:14] <davetoo> Git is... powerful enough to be dangerous

[29-Sep-2011 11:40:22] <dhopp> where are zope roles defined in the ZMI?

[29-Sep-2011 11:40:53] <davetoo> acl_users probably?

[29-Sep-2011 11:40:56] <rocket> Jane_Curry: github.com has many training tutorials actually

[29-Sep-2011 11:41:15] <rocket> Jane_Curry: its actually a VERY helpful resource

[29-Sep-2011 11:41:25] <dhopp> davetoo: that's what I thought..but I created a role there and I don't see it as an option on the groups -> role mappings

[29-Sep-2011 11:42:06] <davetoo> in /zport/acl_users/roleManager/manage_roles  ?

[29-Sep-2011 11:42:16] <davetoo> not /acl_users but /zport/acl_users ?

[29-Sep-2011 11:42:27] <davetoo> I haven't actually tried creating a new role, though.

[29-Sep-2011 11:42:44] <nyeates> Jane_Curry: I agree, there are tons of git resources out there and help.github.com is superb. On the other hand, there is nothing like focused live training - I think I will do this, but to do it with limited scope at first

[29-Sep-2011 11:42:51] <Jane_Curry> dhopp: to see roles, point browser at <zenoss>:8080/zport/manage_access - shows you roles and permissions

[29-Sep-2011 11:42:59] <dhopp> davetoo:  that's where I did it...

[29-Sep-2011 11:43:22] <Jane_Curry> Thanks rocket - I'll go and check

[29-Sep-2011 11:43:28] <davetoo> dunno; I'm looking right now at I can't tell if the ones under zport are inherited from the top level or not.

[29-Sep-2011 11:43:36] <dhopp> Jane_Curry:  I did that and I'm able to assign a permission to the role..but it doesn't show up in a drop down when I try to do an ldap group to role mapping…only Manager,ZenManager and ZenUser show up

[29-Sep-2011 11:43:42] <davetoo> (I'm looking at an Enterprise VM with the ZenOperator pack/role installed)

[29-Sep-2011 11:43:56] <Jane_Curry> If the left hand Acquired box is checked then permissions are inherited - most ARE

[29-Sep-2011 11:44:27] <davetoo> ZenOperator is defined at /zport/acl_users, but does not appear at /acl_users (but the other three/four do [ you don't see Owner in the zenoss app/portall] )

[29-Sep-2011 11:44:38] <Jane_Curry> dhopp: You have refreshed your browser??

[29-Sep-2011 11:44:45] <davetoo> so.. I guess I don't know what else is required

[29-Sep-2011 11:44:56] <dhopp> Jane_Curry:  yes…and restarted zenoss

[29-Sep-2011 11:45:09] <dhopp> actually the role I setup was setup like 3 days ago…I

[29-Sep-2011 11:45:18] <Jane_Curry> Hmmm - restarting zenoss defintely shouldn't be required

[29-Sep-2011 11:45:21] bigegor_ is now known as bigegor

[29-Sep-2011 11:45:53] * davetoo reminds self of that "some day" action item to build an ldap server for dev/test

[29-Sep-2011 11:46:11] <dhopp> Jane_Curry:  yeah I know a zope restart should have been the most I needed but I through the hammer at it :-

[29-Sep-2011 11:46:13] <dhopp> :-)

[29-Sep-2011 11:46:19] <dhopp> *threw

[29-Sep-2011 11:47:26] <Jane_Curry> Can anyone help me with creating a new "My Commands" icon at the bottom of the left-hand menu??????  Also a new "My Commands" at the top right of the list of devices?????

[29-Sep-2011 11:47:28] <Hackman238> Off topic, but wow... --> http://www.wtop.com/?nid=681&sid=2568864

[29-Sep-2011 11:47:54] <jrh0090> @Jane sure one sec

[29-Sep-2011 11:48:20] <Jane_Curry> jrh0090 - if you can do that in 1 sec then you are my hero!

[29-Sep-2011 11:49:41] <jrh0090> in javascript land http://pastie.org/2611792

[29-Sep-2011 11:49:54] <jrh0090> whenever the footer bar is available it just throws that new button on there

[29-Sep-2011 11:50:08] <jrh0090> the '-' means put a little spacer in between

[29-Sep-2011 11:51:18] <Jane_Curry> jhr0090 - wow! Now take me gently with this - I'm not good with javascript...  I guess I put this in a ZenPack browser/resources in a .js file?????

[29-Sep-2011 11:53:13] <jrh0090> Yes, unfortunately we don't have a way to include javascript without the proper zcml though

[29-Sep-2011 11:53:19] <jb> ok, the "Template Path" in 4.x when adding a new template is very confusing.

[29-Sep-2011 11:53:26] <jb> i should just be able to type /Server/Linux, etc.

[29-Sep-2011 11:53:29] <jrh0090> here is an example of declaring a resource and then sending that JS to the client on every request

[29-Sep-2011 11:53:31] <jrh0090> http://pastie.org/2611816

[29-Sep-2011 11:53:55] <dhopp> Jane_Curry:  ok weird…I went back to manage_access and the role wasn't there..I added it and now it's an option to map to a zope role

[29-Sep-2011 11:54:25] <SDuensin> jrh0090, you guys need to take a few days and type your brains into the wiki.  Thanks.  :-)

[29-Sep-2011 11:54:41] <jrh0090> @SDUensin, i would just need 10 minutes

[29-Sep-2011 11:55:13] * SDuensin is trying to make his first ZenPack.  Kinda uphill work.

[29-Sep-2011 11:55:16] <jrh0090> so the drop down to adding an organizer looks competely different than the rest and doesn't facilitate typing ahead if you are used to how we normally print organizer paths

[29-Sep-2011 11:55:52] <Jane_Curry> jrh0090 - so in your second paste, you file with thextra button in is in browser/resources/ZenVMware.js???

[29-Sep-2011 11:56:19] <jrh0090> exactly

[29-Sep-2011 11:56:30] <Jane_Curry> jrh0090 - and the zcml only has to reference the file, not my extension to the menu???

[29-Sep-2011 11:56:32] <jrh0090> you just have to be careful because that JS will be included on every page

[29-Sep-2011 11:56:48] <Jane_Curry> that's what I want - you are indeed my hero!

[29-Sep-2011 11:56:51] <jrh0090> right

[29-Sep-2011 11:57:02] <jrh0090> but it just depends on what you do in the menu

[29-Sep-2011 11:57:16] <Jane_Curry> Now for a curly wurly and a peanut butter sandwich....

[29-Sep-2011 11:57:41] <Jane_Curry> Can you do the extension to have a My Commands a swell as a Commands top-right of the device list???

[29-Sep-2011 11:58:22] <Jane_Curry> SDuensin - have you got my doc on creating ZenPacks????

[29-Sep-2011 11:58:26] <jrh0090> sure, let me look up the component name

[29-Sep-2011 11:58:54] <Jane_Curry> jrh0090 - that's the bit where knowing where to look takes you 2 mins and me 2 hours

[29-Sep-2011 11:59:01] <SDuensin> Jane_Curry, not unless your name used to be Matt Ray.  :-)   That's who's I'm reading.

[29-Sep-2011 11:59:17] <jrh0090> ok so instead of using the footer_bar use the component detail-toolbar

[29-Sep-2011 11:59:18] <jb> rocket: i'm not seeing any documentation about SSH keys/etc for zencommand datasources..

[29-Sep-2011 11:59:35] <jb> rocket: i do see a "Use SSH" box in the "Edit Data Source" dialog for a zencommand data source.

[29-Sep-2011 11:59:38] <jrh0090> generally if you have firebug you can kinda click around kind of findthe names

[29-Sep-2011 11:59:58] <Jane_Curry> Try this then - docs/DOC-10268

[29-Sep-2011 12:00:01] [disconnected at Thu Sep 29 12:00:01 2011]

[29-Sep-2011 12:00:02] [connected at Thu Sep 29 12:00:02 2011]

[29-Sep-2011 12:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[29-Sep-2011 12:00:19] <Jane_Curry> firebug ???

[29-Sep-2011 12:00:23] <SDuensin> Jane_Curry, awesome!  Thank you!

[29-Sep-2011 12:00:26] <rocket> jb you basically set the zproperties on what keys to use

[29-Sep-2011 12:01:01] <Jane_Curry> You're welcome - it's kinda a hard slog first time around - doesn't get much easier second time

[29-Sep-2011 12:01:03] <jrh0090> yeah it is an extension on firefox that really helps with js development

[29-Sep-2011 12:01:09] <SDuensin> Jane_Curry, lol

[29-Sep-2011 12:02:02] <SDuensin> Jane_Curry, sad part is, I make much better sense out of the code than the crazy web UI.  Parts of that thing and I just do not get along.

[29-Sep-2011 12:02:26] <jb> rocket: gotcha.  thanks.

[29-Sep-2011 12:05:56] <nyeates> Jane_Curry: go get firebug now.... no if's and's or buts around it

[29-Sep-2011 12:06:26] <nyeates> amazing in-browser html and DOM and js exploration and editing and testing

[29-Sep-2011 12:06:53] <Hackman238> nyeates: Did you get my verbose PM?

[29-Sep-2011 12:07:53] <nyeates> Hackman238: yeah trying to catch up here :-)

[29-Sep-2011 12:08:10] <Hackman238> nyeates: Alrighty

[29-Sep-2011 12:09:49] <SDuensin> Jane_Curry, this is like actual documentation.  I'm scared.  :-D

[29-Sep-2011 12:15:45] <nyeates> OK, dev session is OVER all....thank jrh0090 for helping out like a madman.

[29-Sep-2011 12:15:57] <nyeates> I cant believe how fast he whipped out js for Jane :-)

[29-Sep-2011 12:15:59] <nyeates> Thanks all

[29-Sep-2011 12:16:10] <ablew> hey all, is it possible to supress event notifications from a host if it's ICMP check fails?

[29-Sep-2011 12:18:03] <dhopp> ablew:  do you have a special ICMP check or do you just mean the standard ping check?

[29-Sep-2011 12:18:10] <ablew> standard ping

[29-Sep-2011 12:18:52] <dhopp> ablew:  it will do that by default if the ping failure comes before the other ones…

[29-Sep-2011 12:19:20] <Jane_Curry> MANY thanks jrh0090 - if you get the second sample it would be much appreciated here

[29-Sep-2011 12:19:30] <dhopp> ablew: if for example ping is successful..device dies..then http check fails you will still see the http failure and then you will see a ping down event the next ping cycle

[29-Sep-2011 12:19:46] <Jane_Curry> nyeates - we have had the recorder running on this haven't we?? Enormously helpful. Thanks

[29-Sep-2011 12:20:26] <dhopp> ablew: but I believe you could do something with an event transformation in that case..you would still see the initial http failure, but you should be able to then suppress that event if a ping failure then happens later

[29-Sep-2011 12:24:57] <ablew> dhopp: my issue is people not putting a device into maintenance mode on reboot, and then the host spams our inboxes with kernel messages related to the reboot when it comes back

[29-Sep-2011 12:26:40] <dhopp> ablew: you mean from syslog?

[29-Sep-2011 12:33:08] <ablew> dhopp: yea

[29-Sep-2011 12:36:18] <nyeates> Jane_Curry: recorder running. I make it a habit now to post this weeks content right before the next IRC meeting, along with summary of what we talked about.

[29-Sep-2011 12:36:48] <nyeates> I dunno how long pasties last for tho :-/

[29-Sep-2011 12:37:06] <jmp242> the pastebins last as long as the submitter selects

[29-Sep-2011 12:37:15] <jmp242> usually 1hr, 1day, 1week 14 days or 1 month

[29-Sep-2011 12:37:21] <jmp242> are the chouices

[29-Sep-2011 12:37:37] <jmp242> I've see 14 days as the default

[29-Sep-2011 13:19:43] <jb> hrm, someone give me a hand with some logic here

[29-Sep-2011 13:21:53] <jb> i have a zencommand currently returns the difference of a value since last poll

[29-Sep-2011 13:22:30] <jb> ie

[29-Sep-2011 13:22:33] <jb> Bind9 is running. 227 successfull requests, 0 referrals, 4 nxdomains since last check. | 'success'=227 'referral'=0 'nxrrset'=32 'nxdomain'=4 'recursion'=0 'failure'=0 'duplicate'=0 'dropped'=0

[29-Sep-2011 13:22:49] <jb> would I set these datapoints as COUNTERs?

[29-Sep-2011 13:26:33] <Hackman238> jb: counters are for rates

[29-Sep-2011 13:26:53] <Hackman238> jb: derrive or absolute I think would make sense. Let me double check

[29-Sep-2011 13:27:49] <Hackman238> jb: My mistake, gauge seems like it would be the choice

[29-Sep-2011 13:27:58] <Hackman238> jb: http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html

[29-Sep-2011 13:28:06] <rmatte> Hackman238: derive is similar to counter only it allows negative values

[29-Sep-2011 13:28:11] <rmatte> I believe that's the only difference

[29-Sep-2011 13:29:49] <Hackman238> rmatte: derive is like counter with no rollover

[29-Sep-2011 13:30:05] <rmatte> right

[29-Sep-2011 13:30:08] <Hackman238> rmatte: plus negatives

[29-Sep-2011 13:30:25] <Hackman238> rmatte: I think gauge would be best, dont you?

[29-Sep-2011 13:30:37] <rmatte> hmmm, that reminds me, still need to write up a guide for adding interface reset detection via ifLastChange polling

[29-Sep-2011 13:30:57] <Hackman238> rmatte: sounds cool

[29-Sep-2011 13:31:06] <rmatte> since he's graphing a straight difference, yes gauge is the best choice

[29-Sep-2011 13:31:55] <rmatte> The guide I want to write is effectively an addon to this: docs/DOC-2494

[29-Sep-2011 13:32:01] <rmatte> I may actually just add it right in to that guide

[29-Sep-2011 13:34:00] <Hackman238> rmatte: Yeah, that'll be interestign to read

[29-Sep-2011 13:34:57] <rmatte> god jive is a pain in the ass to work with

[29-Sep-2011 13:36:22] <Hackman238> rmatte: Indeed it is

[29-Sep-2011 13:37:07] <Guest97443> Hi all , is there option to display the snmpwalk result which are characters. like for example users signed in a device. or be any result tht are be character. Is zenoss has option for the same? I am not sure whether this is right place to ask such questions. as I was wondering if not, then is this might be something to considered for future?

[29-Sep-2011 13:38:28] <Hackman238> Guest97443: You'll need to write a custom parser to use an snmp data source in this way or create a script that ouptputs nagios format results to a command data source

[29-Sep-2011 13:39:35] <Guest97443> and we can display it in a tab?

[29-Sep-2011 13:40:11] <jb> hmm

[29-Sep-2011 13:40:25] <jb> im wondering how i can turn this in to "per second"

[29-Sep-2011 13:41:05] <Hackman238> jb: the script that populates snmp on the source device can diff the new and old value

[29-Sep-2011 13:41:18] <Hackman238> jb: That'll give you absolute rate

[29-Sep-2011 13:42:09] <Hackman238> jb: If you dont, derive with min=0 might make sense too

[29-Sep-2011 13:42:36] <Hackman238> jb: That'll give you a rate automatically. Either way, the number will be normalized so you might get decimals

[29-Sep-2011 13:43:43] <Hackman238> jb: definately more than a few ways to do it

[29-Sep-2011 13:43:52] <jb> well, heres what I'm doing..

[29-Sep-2011 13:43:55] <jb> i am actually taking a difference

[29-Sep-2011 13:44:01] <jb> http://www.matejunkie.com/update-12-bind-check-plugin-for-nagios/

[29-Sep-2011 13:44:07] <Guest97443> Hackman: speaking of command data source. Is it possible to graph the result from command data source. Coz I tried once with a result but it didnt create any data point which can be mapped to graph.

[29-Sep-2011 13:44:10] <Hackman238> service zends status

[29-Sep-2011 13:44:12] <Hackman238> whoops

[29-Sep-2011 13:44:14] <Hackman238> Sorry

[29-Sep-2011 13:46:00] <Hackman238> Guest97443: Yes. You'll need to make sure the command datasource output is parsable. stick with nagios format (example: OK|myDataPoint=55)

[29-Sep-2011 13:46:37] <Hackman238> Guest97443: Once the output is good you'll need to add a datapoint with the same name as one of the parsable members of the output (example: myDataPoint)

[29-Sep-2011 13:46:46] <Hackman238> Guest97443: You can then graph myDataPoint

[29-Sep-2011 13:47:19] <Guest97443> for example : /usr/local/zenoss/common/bin/snmpwalk -cpublic -v2c "device ip" "oid"

[29-Sep-2011 13:47:22] <Guest97443> i used this

[29-Sep-2011 13:47:39] <Guest97443> to graph the result and i can reduce my polling time

[29-Sep-2011 13:48:01] <Guest97443> is it ok? or any mistake in the format?

[29-Sep-2011 13:48:51] <jb> Hackman238: i'm not sure what the benefit of this script taking the difference of two polls so close to each other..

[29-Sep-2011 13:49:18] <Hackman238> Guest97443: That format works for the command, but the output of the command will not be readable to zenoss. You'll need to chop it up with a script to look like: OK|myDataPonit=55

[29-Sep-2011 13:49:39] <rmatte> Hackman238: voila docs/DOC-2494

[29-Sep-2011 13:49:43] <Hackman238> Guest97443: Alternatively you can use smnp datapoints for OIDs that return numerical data only.

[29-Sep-2011 13:50:06] <Guest97443> main issue was due to weird reason i cudnt add the oid directly in the zenoss and parse it for graphing. when i do snmpwalk i get results from gui and from command line

[29-Sep-2011 13:50:20] <Hackman238> Guest97443: If the polling cycle time for zenperfsnmp is too great, you might want to consider distriuted collectors or adding extra logical collectors for more 'copies' of zenperfsnmp

[29-Sep-2011 13:50:46] <Guest97443> message/61602#61602

[29-Sep-2011 13:50:51] <Hackman238> Guest97443: For OIDs in zenoss, you must input as all numerical and sometimes add a .0 to the end

[29-Sep-2011 13:51:12] <rmatte> hmmm, just noticed one small error in a step...

[29-Sep-2011 13:51:24] <Hackman238> Guest97443: Ah. I'm Shane Can you post the ouptut?

[29-Sep-2011 13:51:43] <jbk> if it's not a value in a table you want to fetch, it needs .0

[29-Sep-2011 13:51:46] <Guest97443> Hackman238: here u can see it. but to security reason my company dont me to realive the device oid and name so it became hard for me to discuss in detail abt the device name

[29-Sep-2011 13:52:02] <jbk> btw, are there any plans to leverage SNMP mibs more than just trap info?

[29-Sep-2011 13:52:04] <Hackman238> Guest97443: Thats ok. What type of device?

[29-Sep-2011 13:52:07] <rmatte> here, fixed

[29-Sep-2011 13:52:09] <rmatte> there*

[29-Sep-2011 13:52:21] <Guest97443> its a network device i think its kinda tap

[29-Sep-2011 13:52:28] <Hackman238> jbk: for creating templates from MIBs?

[29-Sep-2011 13:52:35] <Hackman238> jbk: Not that I'm aware of.

[29-Sep-2011 13:52:41] <jbk> templates

[29-Sep-2011 13:52:44] <jbk> key values

[29-Sep-2011 13:52:46] <jbk> etc.

[29-Sep-2011 13:52:58] <Hackman238> jbk: Not that I'm aware of.

[29-Sep-2011 13:53:16] <jbk> that sucks

[29-Sep-2011 13:53:27] <Hackman238> jbk: Its a bit contrary to the zenoss templating system

[29-Sep-2011 13:53:29] <Guest97443> i tried with .0 and without .0 tricks but no luck and i tried to index the results and all but still no luck

[29-Sep-2011 13:54:00] <jbk> to have to put in .1.2.3.4.5.6.7.8.0 instead of "enterprises.foo.bar" for the oid?

[29-Sep-2011 13:54:16] <Hackman238> Guest97443: A tap? Packet inspection device?

[29-Sep-2011 13:54:20] <Guest97443> iHackman238: i can see in gui rrd files r been used but in the linux box i dont see the files such created

[29-Sep-2011 13:54:30] <Guest97443> ya kinda

[29-Sep-2011 13:54:38] <Hackman238> jbk: No no, auto creation of templates from MIBs

[29-Sep-2011 13:55:02] <Hackman238> Guest97443: whats the OID?

[29-Sep-2011 13:55:23] <jbk> oh that i can see

[29-Sep-2011 13:55:26] <jbk> that would be a bit silly

[29-Sep-2011 13:55:33] <Hackman238> jbk: Right.

[29-Sep-2011 13:55:39] <Hackman238> jbk: Thats what I thought you'd meant.

[29-Sep-2011 13:55:57] <Hackman238> jbk: change in supported OID lexicon might be possible.

[29-Sep-2011 13:56:05] <Guest97443> Hackman238: sorry mate I m been restricted to give tht...

[29-Sep-2011 13:56:12] <rmatte> "We are pleased to announce that Nova Networks is now a recognized Cisco Advanced Managed Services Channel Partner." - Fastastic, we busted our asses for that lol

[29-Sep-2011 13:56:27] <Hackman238> rmatte: LOL

[29-Sep-2011 13:56:48] <jb> ugh i suck at this

[29-Sep-2011 13:56:51] <jb> ok. soo.

[29-Sep-2011 13:57:04] <jb> this script takes two polls back to back, then subtracts those two values

[29-Sep-2011 13:57:06] <Hackman238> Guest97443: can you post the output of a snmpget of the oid santiized of device names, etc?

[29-Sep-2011 13:57:18] <Guest97443> sure

[29-Sep-2011 13:57:24] <jb> so, zenoss calls this script every 5 minutes via zencommand.

[29-Sep-2011 13:57:33] <jb> trying to figure out what kind of data this is giving me

[29-Sep-2011 13:59:27] <Guest97443> HAckman238:[zenoss@zenoss root]$ snmpget -cpublic -v2c "device IP" "oid without.0" SNMPv2-SMI::enterprises."oid without.0" = No Such Instance currently exists at this OID

[29-Sep-2011 13:59:28] <Hackman238> jb: You'll want to have y=d(t)-d(t-1)

[29-Sep-2011 14:00:09] <Hackman238> jb: the delta equals the last data point value unadjusted minute the previos data point value unadjusted

[29-Sep-2011 14:00:22] <Hackman238> *s/minute/minus/

[29-Sep-2011 14:00:29] <Hackman238> Sorry....very tired. Been a long day

[29-Sep-2011 14:00:34] <rmatte> Hackman238: do you guys actively poll for interface status or use traps?

[29-Sep-2011 14:00:35] <rocket> jb: why are you grabbing the data twice?

[29-Sep-2011 14:00:47] <Guest97443> Hackman238: [zenoss@zenoss root]$ snmpget -cpublic -v2c "device IP" "oid without.0".0 SNMPv2-SMI::enterprises."oid with.0".0 = No Such Instance currently exists at this OID

[29-Sep-2011 14:01:06] <jb> rocket: well, bind outputs a continuous number.

[29-Sep-2011 14:01:07] <jb> ie

[29-Sep-2011 14:01:18] <Hackman238> rmatte: traps only

[29-Sep-2011 14:01:23] <jb> one poll gives you 22518

[29-Sep-2011 14:01:31] <Hackman238> rmatte: Way too much extra load with no value to poll for the status

[29-Sep-2011 14:01:32] <jb> the next, 22293

[29-Sep-2011 14:01:39] <rmatte> Guest97443: don't do it that way, you can't use the translated OIDs like that

[29-Sep-2011 14:01:45] <rmatte> Guest97443: add -On to the snmpwalk command

[29-Sep-2011 14:01:53] <Hackman238> Guest97443: You'

[29-Sep-2011 14:01:56] <rmatte> Guest97443: and it'll output them as actual OIDs instead

[29-Sep-2011 14:01:58] <jb> rocket: so the nagios script is taking the difference of those two numbers and returning it

[29-Sep-2011 14:02:00] <Hackman238> ll need to use numerical OIDs

[29-Sep-2011 14:02:10] <rocket> jb I understand that but why do you need the difference?

[29-Sep-2011 14:02:22] <rmatte> Hackman238: ah, I find polling for status is much more effective even though it does create extra load

[29-Sep-2011 14:02:26] <jb> well, it's providing a rate of change, I suppose.

[29-Sep-2011 14:02:30] <jb> this is from an example script I found.

[29-Sep-2011 14:02:50] <rocket> jb: well the rate of change would be taken care of via the rrd type

[29-Sep-2011 14:02:56] <Hackman238> rmatte: we really only setup up/down traps for important ports anyway

[29-Sep-2011 14:02:59] <Guest97443> [zenoss@zenoss root]$ snmpwalk -cpublic -v2c "device IP" "oid without.0" SNMPv2-SMI::enterprises."oid without.0".1.23839 = INTEGER: 10620 SNMPv2-SMI::enterprises."oid without.0".1.23840 = INTEGER: 10609 SNMPv2-SMI::enterprises."oid without.0".1.23841 = INTEGER: 10677 SNMPv2-SMI::enterprises."oid without.0".1.23842 = INTEGER: 10698 SNMPv2-SMI::enterprises."oid without.0".1.23843 = INTEGER: 10705 SNMPv2-SMI::enterprises."oid without.0

[29-Sep-2011 14:03:08] <jb> rocket: ok, so is there any way to get per/second query information?

[29-Sep-2011 14:03:10] <rmatte> Hackman238: traps can go MIA on their way to Zenoss

[29-Sep-2011 14:03:10] <rocket> derive for example would be the rate of change between the last poll and the new poll

[29-Sep-2011 14:03:17] <jb> ok, got it.

[29-Sep-2011 14:03:21] <Guest97443> as u see i get results for snmpwalk not snmpget

[29-Sep-2011 14:03:24] <rocket> jb just use derive

[29-Sep-2011 14:03:24] <Guest97443> oh i will try tht

[29-Sep-2011 14:03:31] <Hackman238> rmatte: There would be a hundred thousand events of up/downs if we did accross the board.

[29-Sep-2011 14:03:47] <rmatte> Hackman238: lol

[29-Sep-2011 14:03:50] <Hackman238> rmatte: thats true, but then its not a zenoss problem.

[29-Sep-2011 14:03:51] <rocket> its  really ( dp2 - dp1 ) / 5min polling cycle

[29-Sep-2011 14:04:04] <jb> ok, and use RPN to get per second?

[29-Sep-2011 14:04:10] <Hackman238> jb: derive and set min=0

[29-Sep-2011 14:04:13] <rocket> no need

[29-Sep-2011 14:04:36] <rmatte> Hackman238: If you're monitoring server ports then yeh, we only monitor WAN/uplink ports

[29-Sep-2011 14:04:42] <rocket> what you are doing is ( rate2 - rate1) / time

[29-Sep-2011 14:04:59] <jb> time defined by the polling interval..

[29-Sep-2011 14:05:03] <rocket> which in most cases doesnt mean anything

[29-Sep-2011 14:05:06] <rocket> yes

[29-Sep-2011 14:05:26] <Hackman238> rmatte: Unfortuntely even just uplinks and important ports are up in the many thousands

[29-Sep-2011 14:05:47] <jb> correct.. sorry if i'm misunderstanding this.  But, I would prefer to have "queries per seconds" stat, not "queries per 5 minutes"

[29-Sep-2011 14:05:49] <jb> possible?

[29-Sep-2011 14:06:10] <rmatte> Hackman238: right, but they shouldn't be resetting often

[29-Sep-2011 14:06:29] <Hackman238> rmatte: You'd be suprised

[29-Sep-2011 14:06:32] <Hackman238> rmatte: I know I was

[29-Sep-2011 14:06:34] <Hackman238> LOL

[29-Sep-2011 14:06:42] <rmatte> I wish custom properties applied to components instead of just to devices

[29-Sep-2011 14:06:50] <rmatte> then you could create something like cWanInterface

[29-Sep-2011 14:06:54] <rmatte> and set it to No or Yes

[29-Sep-2011 14:06:59] <rocket> well if you are sure you want that exactly to be that .. do your double polling with 1 second between and then use counter

[29-Sep-2011 14:07:00] <rmatte> then use a transform based on that

[29-Sep-2011 14:07:04] <rmatte> and drop events selectively

[29-Sep-2011 14:07:08] <Hackman238> rmatte: Polling or traps mostly just pref.

[29-Sep-2011 14:07:23] <Guest97443> ray: oh tht shows the entire oid... unfortunately i m restricted to provide oid or device name... i know its hard without tht.... but as u can see the results of snmpwalk i get the integer values.

[29-Sep-2011 14:07:29] <rocket> then it will save the raw result

[29-Sep-2011 14:07:32] <rmatte> The polling works well for us because we have really anal customers who will come back for every little thing that we miss

[29-Sep-2011 14:07:37] <Hackman238> rmatte: Thats a problem too

[29-Sep-2011 14:07:40] <rocket> and not divide by time a second time

[29-Sep-2011 14:07:43] <rmatte> with both ifOperStatus and ifLastChange polling we don't miss anything anymore

[29-Sep-2011 14:07:53] <Hackman238> rmatte: Some of our transforms are pages lnog to account for scenarios, cases, etc

[29-Sep-2011 14:08:00] <rmatte> yeh

[29-Sep-2011 14:08:13] <Hackman238> rmatte: results in a ton of extra cycles.

[29-Sep-2011 14:08:24] <rmatte> I try to keep my transforms as compact as possible

[29-Sep-2011 14:08:24] <Hackman238> rmatte: The other big reason is we use splunk

[29-Sep-2011 14:08:38] <rmatte> we just started testing splunk for one of our clients

[29-Sep-2011 14:08:48] <Hackman238> rmatte: syslog and traps get dumped in splunk which is very easy to search quickly

[29-Sep-2011 14:09:43] <rmatte> cool

[29-Sep-2011 14:10:08] <Hackman238> rmatte: Our Zenoss rarely finds something down before staff do. Zenoss is more for metrics, capacity analysis/planning, and casual monitoring of the network

[29-Sep-2011 14:10:18] <rmatte> ah

[29-Sep-2011 14:10:31] <rmatte> our whole game is alerting customers, so we have to notice it's down

[29-Sep-2011 14:10:31] <rmatte> lol

[29-Sep-2011 14:10:34] <Hackman238> rmatte: NOCs are staffed 24/7/365 everywhere

[29-Sep-2011 14:10:44] <Hackman238> rmatte: Gotcha

[29-Sep-2011 14:10:58] <rmatte> so is ours, but our guys just watch the ticketing system, and Zenoss feeds in to that (which I believe is what you guys do too)

[29-Sep-2011 14:11:13] <rmatte> It's really the only way to go these days

[29-Sep-2011 14:11:19] <Hackman238> rmatte: We try to not have the customer see any problems

[29-Sep-2011 14:11:30] <rmatte> some of our competitors in this city have people sitting there staring at screens all day and generating tickets by hand

[29-Sep-2011 14:11:33] <Hackman238> rmatte: Definately. Automatic ticketing is boss

[29-Sep-2011 14:11:34] <rmatte> makes me cry to think about

[29-Sep-2011 14:11:43] <Hackman238> rmatte: Or laugh

[29-Sep-2011 14:11:52] <rmatte> our guys have trouble keeping up with the auto-tickets sometimes, makes me wonder how many alerts they miss when doing it manually

[29-Sep-2011 14:12:09] <Hackman238> rmatte: I hear that!

[29-Sep-2011 14:12:11] <rmatte> "It's all red!"

[29-Sep-2011 14:12:21] <Hackman238> rmatte: "Red alert! Update the resume!"

[29-Sep-2011 14:12:24] <jbk> heh

[29-Sep-2011 14:12:29] <jbk> at a previous job

[29-Sep-2011 14:12:35] <jbk> we lost power in one of our datacenters

[29-Sep-2011 14:12:54] <jbk> this also happened to be where all teh dedicated circuits for our field locations (hundreds of them) terminated

[29-Sep-2011 14:13:04] <Hackman238> jbk: yike!

[29-Sep-2011 14:13:07] <jbk> so there was an entire rack of csu/dsus all in red alarm

[29-Sep-2011 14:13:17] <rmatte> brutal

[29-Sep-2011 14:13:18] <dhopp> rmatte: but do you create a ticket for every event?  Or do you have logic for which events require a ticket to be opened?

[29-Sep-2011 14:13:19] <Hackman238> jbk: spells doom for the company quarterly losses LOL

[29-Sep-2011 14:13:25] <jbk> the network guy was like 'i'm not even sure how to call this in to sbc (local loop provider)'

[29-Sep-2011 14:13:30] <jb> rocket: ok, so, the only real way to get "per second," is to double poll.

[29-Sep-2011 14:13:31] <jbk> naah

[29-Sep-2011 14:13:36] <jbk> this happened in the middle of the night

[29-Sep-2011 14:13:53] <jbk> but i told him 'when they ask for the circuit id, just say 'yes''

[29-Sep-2011 14:14:01] <Hackman238> jbk: LOL

[29-Sep-2011 14:14:14] <rmatte> dhopp: every unique event generates a ticket, events that are the same as tickets that are already open get correlated back to those tickets (there is an event count value in each ticket and first time/last time values that get automatically updated.

[29-Sep-2011 14:14:17] <jbk> thankfully it was mostly windows servers in there

[29-Sep-2011 14:14:35] <jbk> our main internal mail relay + another small web server were the only unix boxes there

[29-Sep-2011 14:14:46] <Hackman238> dhopp: Logic is a must otherwise the CMS fold would be pissed!

[29-Sep-2011 14:14:50] <Hackman238>

[29-Sep-2011 14:15:01] <jbk> so i just had to verify they were ok (it wasn't enough for them that I could reach them via the internet)

[29-Sep-2011 14:15:05] <jbk> then go home

[29-Sep-2011 14:15:12] <rmatte> dhopp: we can also place tickets in a "pending reminder" state, which will automatically move them to open if the event re-occurs

[29-Sep-2011 14:15:15] <Hackman238> jbk: Crazy.

[29-Sep-2011 14:15:28] <rmatte> and we have a "suppressed" state that prevents that for chronic events such as an interface being down

[29-Sep-2011 14:15:37] <rocket> Hackman238: you guys might really like the new correlation tools if you guys end up purchasing them

[29-Sep-2011 14:16:04] <Hackman238> rocket: impact?

[29-Sep-2011 14:16:10] <dhopp> Hackman238:  I ask because our current monitoring system opens a ticket for everything (there is a history of the event AND there is a ticket opened).  So if CPU jumps to 80% for 5 minutes a ticket is opened and then the alert will clear 3 minutes later and causes a lot of tickets being closed with 'Temporary alert'...

[29-Sep-2011 14:16:10] <rocket> yeah

[29-Sep-2011 14:16:19] <Hackman238> rocket: maybe when we take over cloud.

[29-Sep-2011 14:16:41] <rocket> impact will lump tickets together based on whats going on

[29-Sep-2011 14:16:44] <Hackman238> rocket: I'm not sure we could jsutify impact for the network

[29-Sep-2011 14:16:58] <rmatte> dhopp: oh, I have rules in the ticket create script for certain thresholds (such as CPU) which only generate a ticket after 3 consecutive events (event count of 3)

[29-Sep-2011 14:16:59] <Hackman238> rocket: Our ticketing system does that already

[29-Sep-2011 14:17:06] <rmatte> dhopp: which eliminates a lot of the false positives

[29-Sep-2011 14:17:23] <rmatte> dhopp: and if a ticket is already open it doesn't generate a new ticket anyways, only if the original ticket is closed

[29-Sep-2011 14:17:25] <Hackman238> rocket: unless its less costly that I thought

[29-Sep-2011 14:17:31] <dhopp> Hackman238:  I personally think that a critical event (device actually down) should open a ticket automatically but other events (maybe disk space) might not need that kind of tracking (since there is still a history in Zenoss)

[29-Sep-2011 14:17:32] <Hackman238> *than I thought

[29-Sep-2011 14:17:40] <rocket> Hackman238: I have no idea on its current cost

[29-Sep-2011 14:18:01] <jb> rocket: *smirk* .. DERIVE = 2nd poll - 1st poll / 300 seconds.

[29-Sep-2011 14:18:07] <jb> rocket: which would actually be "per second"

[29-Sep-2011 14:18:13] <jb> well, an average

[29-Sep-2011 14:18:17] <Hackman238> dhopp: device downs check the CMS to see if the client was terminated, suspended, in maint, etc. before doing tickets on crits

[29-Sep-2011 14:18:35] <rmatte> dhopp: the thing with us is that we have 15 Zenoss servers, soon to be 17, and they need to all feed alerts back to a single location where we can manage the alerts from

[29-Sep-2011 14:18:42] <jb> rocket: i was reading docs/DOC-4426

[29-Sep-2011 14:18:56] <rocket> jb yes as I was trying to explain .. they both basically do the same thing

[29-Sep-2011 14:19:07] <rocket> my recomendation is less complicated

[29-Sep-2011 14:19:10] <jb> sorry, I was super confused

[29-Sep-2011 14:19:20] <dhopp> rmatte: ah…we will have multiple Zenoss servers but they are all one system so we already have the central location..

[29-Sep-2011 14:19:25] <Hackman238> jb: values will still be normalized though

[29-Sep-2011 14:19:30] <Hackman238> jb: so you could get decimals

[29-Sep-2011 14:19:32] <jb> so, I don't need to do anything for per second.  just use DERIVE.

[29-Sep-2011 14:19:36] <jb> rohjy

[29-Sep-2011 14:19:38] <jb> right

[29-Sep-2011 14:19:51] <Hackman238> jb: jsut cut off the decimals..close enough

[29-Sep-2011 14:19:55] <jb> since DERIVE divides the difference by the interval.

[29-Sep-2011 14:19:59] <jb> blah confusing

[29-Sep-2011 14:20:26] <rmatte> dhopp: ours are all separate instances, some of our clients have dedicated servers, and we also have multi-client servers

[29-Sep-2011 14:20:58] <dhopp> Hackman238:  yeah that would be nice..but what happens in our current monitoring system is we get CPU > 80% open ticket, cpu alert cleared…CPU > 80% open ticket, cpu alert cleared…rinse and repeat…and then the tickets have to be manually closed (because we 'should' be putting notes as to why that happened)…

[29-Sep-2011 14:21:29] <Hackman238> dhopp: Ah gotcha.

[29-Sep-2011 14:21:40] <dhopp> so one of the things I got beat up on when demoing Zenoss was "where is the incident/ticket number"..well Zenoss isn't a ticketing system..it's a monitoring/alert system we can tie it into a ticketing system if needed but we need to do it intelligently

[29-Sep-2011 14:22:08] <rmatte> dhopp: If a chronic issue like that occurs with us, the first ticket is left open, meaning the correlation comes in to play and there is no other tickets opened... then our NOC guys contact the engineer responsible for the account to have them investigate the issue.  If an increase in threshold value is required then it's sent to me and I make the adjustment.

[29-Sep-2011 14:22:42] <rmatte> We also only generate tickets for CPU if it's been over the threshold for 3 consecutive cycles

[29-Sep-2011 14:22:43] <Hackman238> rmatte: Yep, thats the way to do it

[29-Sep-2011 14:22:53] <rmatte> we're not interested in spikes

[29-Sep-2011 14:22:59] <rmatte> we're interested in the system being pinned

[29-Sep-2011 14:23:07] <Hackman238> rmatte: Meh...less its a core network device or an edge, we dont care about CPU

[29-Sep-2011 14:23:36] <rmatte> we monitor CPU on basically everything since it's a requirement for us

[29-Sep-2011 14:23:45] <dhopp> rmatte: the other thing I'm struggling with is the ticket that gets opened by our monitoring system is NOT in the same system that rest of our tasks/tickets go in..so it's fricken mess…the NOC will contact an engineer if it's a chronic problem but there is no logic when a ticket is opened so if the alert clears and then reappears a new ticket is open...

[29-Sep-2011 14:23:48] <rmatte> our customers really only pay us to monitor their more critical devices in the first place

[29-Sep-2011 14:23:51] <Guest97443> ticketing system which one you guys use? how is it with remedy or service center or so?

[29-Sep-2011 14:24:07] <dhopp> so I'm curious..how do you correlate the event in Zenoss to an external ticket system

[29-Sep-2011 14:24:09] <rmatte> Guest97443: we use OTRS, but a heavily customized version of it

[29-Sep-2011 14:24:15] <Hackman238> dhopp: yikes

[29-Sep-2011 14:24:29] <rmatte> we developed a lot of customizations and the integration between Zenoss and OTRS

[29-Sep-2011 14:24:47] <Hackman238> Guest97443: We have a custom CMS written in python called CORE

[29-Sep-2011 14:24:50] <rmatte> we have a perl script that handles the ticket creation (using soap), and I wrote a Zenoss daemon that handles calling the script

[29-Sep-2011 14:25:08] <rmatte> when tickets are successfully created the daemon automatically acknowledges the event in Zenoss

[29-Sep-2011 14:25:11] <dhopp> Hackman238:  yeah it's a mess but it's been that way for so long that people that have been here for years are used to it and that's what they want..regardless if they know it's stupid or not

[29-Sep-2011 14:25:16] <Guest97443> ohh ok... coz I am thinking of remedy... but its in initial phase of discussion

[29-Sep-2011 14:25:17] <rmatte> so you can easily tell which events generated tickets and which didn't

[29-Sep-2011 14:25:31] <rmatte> remedy was too expensive

[29-Sep-2011 14:25:35] <rmatte> and OTRS is every bit as good

[29-Sep-2011 14:25:39] <rmatte> I've used Remedy in the past

[29-Sep-2011 14:26:19] <Guest97443> oh cool then i should test it with zenoss as it coming with strong recoommnendatiaon

[29-Sep-2011 14:26:40] <rmatte> well, the integration we made is very custom and took a long while to develop (since we used an actual API)

[29-Sep-2011 14:26:46] <rmatte> you'd have to do the same

[29-Sep-2011 14:27:02] <rmatte> but then again you'd also have to do it with remedy

[29-Sep-2011 14:27:17] <rmatte> unless you bought enterprise which there's a pack for it

[29-Sep-2011 14:27:31] <Guest97443> ya i used AR remedy in the past... as u said its quite expensive so looking for alternative option... it was intergrated with HP openview

[29-Sep-2011 14:27:45] <Hackman238> You'll all be happy to know the API in v4 is much better

[29-Sep-2011 14:27:54] <rmatte> that's the exact setup that we used in the past, Remedy and HP openview

[29-Sep-2011 14:27:56] <Hackman238> Code it cleaner too

[29-Sep-2011 14:28:01] <rmatte> (at a company I worked for back in the day)

[29-Sep-2011 14:28:10] <Guest97443> lol me too

[29-Sep-2011 14:28:11] <rmatte> but it was insanely expensive

[29-Sep-2011 14:28:12] <rmatte> lol

[29-Sep-2011 14:28:13] <Hackman238> *ears burn from hearing HP OpenView*

[29-Sep-2011 14:28:22] <Guest97443> ha ha

[29-Sep-2011 14:28:23] <rmatte> it was also some archaic version of openview

[29-Sep-2011 14:28:35] <Hackman238> rmatte: the newest version is archais

[29-Sep-2011 14:28:36] <tvincent> HPOV, a consultants dream

[29-Sep-2011 14:28:36] <rmatte> had to forward x window sessions to view network maps

[29-Sep-2011 14:28:41] <rmatte> and it was ugle as all hell

[29-Sep-2011 14:28:43] <rmatte> ugly*

[29-Sep-2011 14:29:01] <Hackman238> tvincent: I'm converting a HPOV to Zenoss now...its madness!

[29-Sep-2011 14:29:02] <Guest97443> yaa true

[29-Sep-2011 14:29:02] <rmatte> I haven't seen the newer versions of it

[29-Sep-2011 14:29:15] <dhopp> Hackman238:  so I have been trying to get people to buy into putting logic that says "these type of events need to have ticket open"…we have our NOC guys with enough power to restart some daemons (tomcat) which often will clear an alert but because so many tickets are opened things sometimes get lost and so a certain event has to reoccur often before somebody finally goes "Why in God's name is this happening so often"?

[29-Sep-2011 14:29:40] <rmatte> HPOV, their slogan should be "After a week of using HPOV you'll be thinking to yourself, 'it shouldn't have to be this hard'"

[29-Sep-2011 14:30:20] <Hackman238> dhopp: Not easy to change people set in their ways LOL

[29-Sep-2011 14:30:40] <Hackman238> rmatte: HPOV, a good reason to hit the bar after work!

[29-Sep-2011 14:30:46] <Hackman238>

[29-Sep-2011 14:30:53] <Guest97443> lol

[29-Sep-2011 14:30:54] <rmatte> the one command I remember using the most with HPOV was "ovtopodump"

[29-Sep-2011 14:31:00] <rmatte> since I was analyst at the time

[29-Sep-2011 14:31:14] <rmatte> Hackman238: haha

[29-Sep-2011 14:31:46] <Hackman238> rmatte: HPOV, cures 'heavy wallet' syndrome everytime!

[29-Sep-2011 14:31:52] <dhopp> rmatte / Hackman238: but HPOV doesn't need customizations..it does so much out of the box :-P

[29-Sep-2011 14:31:53] <rmatte> yup

[29-Sep-2011 14:32:04] <Hackman238> dhopp: LOL

[29-Sep-2011 14:32:04] <tvincent> When you have integrated a external system with a ticketing system. Do you update the tickets constantly with the number of times a event has occured?

[29-Sep-2011 14:32:04] <dhopp>  (an argument I here almost daily with pretty much any commercial software)

[29-Sep-2011 14:32:36] <Guest97443> but wht abt netflow or sflow tools?? i came across scrutinizer... i was doing some testing with it... but i was speaking with them... they told tht they didnt release much latest zenpack for 3.0.3 and above...

[29-Sep-2011 14:32:42] <rmatte> I always love that when people start using Zenoss and whine about the fact that they have to customize it.  "Such and such product works out of the box!", news flash, none of them do.

[29-Sep-2011 14:33:04] <tvincent> dhopp: commercial software isn't bad. Friend at HP told me HPOV had the largest device support of any software out there. Plus it can support millions of devices.

[29-Sep-2011 14:33:05] <rmatte> "Why can't I just put MIBs in and have it automagically monitor everything?", doesn't work that way

[29-Sep-2011 14:33:36] <dhopp> tvincent:  it can…but NO monitoring software can support EVERY environment out of the box

[29-Sep-2011 14:34:07] <Hackman238> tvincent: We do not.

[29-Sep-2011 14:34:10] <tvincent> dhopp: agreed

[29-Sep-2011 14:34:14] <rmatte> tvincent: IBM is pushing us to use Tivoli Service Desk for a monitoring contract that we're getting through them, so now we have to jerry rig up some integration between that and our ticketing system

[29-Sep-2011 14:34:14] <Guest97443> thts true... tweaking for ur requirement is necessity

[29-Sep-2011 14:34:20] <dhopp> tvincent:  so sure I can tell it to monitor devices 1 2 3 and it will get me cpu/memory etc…but it has no idea how device 1 2 3 interact

[29-Sep-2011 14:34:45] <Hackman238> AAHHH!!

[29-Sep-2011 14:34:52] <dhopp> tvincent:  my point was just because it does all this out of the box doesn't mean we don't have to customize it..just means we have to customize it differently

[29-Sep-2011 14:34:58] <Hackman238> *searing pain from sound of Tivoli*

[29-Sep-2011 14:35:00] <tvincent> IBM push IBM products. Go figure.

[29-Sep-2011 14:35:14] <rmatte> tvincent: yeh, who woulda thunk it

[29-Sep-2011 14:35:19] <tvincent> I haven't found TIvoli that bad over the years. Just complete and total overkill in most situations.

[29-Sep-2011 14:35:50] <rmatte> Right now the bane of my existence is Solarwinds

[29-Sep-2011 14:35:51] <jbk> tivoli? stab your eyes out with a spoon now

[29-Sep-2011 14:35:55] <tvincent> rmatte: AMEN!

[29-Sep-2011 14:36:10] <rmatte> and the only reason we're using it is because it provides all these fancy top 10 type graphs and tables that managers love

[29-Sep-2011 14:36:25] <tvincent> rmatte: Custom report in ZenOSS?

[29-Sep-2011 14:36:26] <tvincent> =)

[29-Sep-2011 14:36:27] <Hackman238> ZenOSS CORE + Solutions Architect = Solution; Tivoli/HPOV + Solutions Architect = Madness and Investment Loss

[29-Sep-2011 14:36:40] <rmatte> tvincent: nah, they want an actual dashboard, I'd have to code a wackload of portlets

[29-Sep-2011 14:36:57] <tvincent> Tivoli really shines when it is integrated with all the different tivoli tools.

[29-Sep-2011 14:37:15] <Hackman238> tvincent: not worth the investment. Not a good return on investment.

[29-Sep-2011 14:37:18] <tvincent> Even then that is only really of value if your environment has i-series, and mainframes.

[29-Sep-2011 14:37:44] <rmatte> tvincent: which is exactly why they are pushing it on us, because we'll be supporting i-series and mainframes

[29-Sep-2011 14:37:45] <rmatte> lol

[29-Sep-2011 14:37:48] <rmatte> AS/400

[29-Sep-2011 14:37:57] <tvincent> My last shop was purchasing new ones.

[29-Sep-2011 14:38:14] <tvincent> implimenting a new ERP system on it.

[29-Sep-2011 14:38:29] <tvincent> THis was also the place that had millions of lines of COBOL on HP 3000's.

[29-Sep-2011 14:38:33] <tvincent> No zenpack for that. =)

[29-Sep-2011 14:38:46] <dhopp> tvincent:  but there could be one :-P

[29-Sep-2011 14:39:06] <Hackman238> tvincent: Its doable for a low price

[29-Sep-2011 14:39:07] <tvincent> dhopp: No, the things were so damn old I was afraid to breathe on them.

[29-Sep-2011 14:39:31] <gwb2353> we have a z10 mainframe used for ERP here... would love to have a zenpack to allow us easy trending (yes, there are native tools that do this, but I like consolidation)

[29-Sep-2011 14:40:37] <Hackman238> Packs for exotic and archane devices are uncommon because the devices to test with are...well...uncommon

[29-Sep-2011 14:41:04] <rmatte> yup

[29-Sep-2011 14:41:07] <gwb2353> how did you get the F5 zenpack together?  it's very nice.  Was that sponsored by F5 or a client?

[29-Sep-2011 14:41:55] <gwb2353> ..and huge kudos to whomever @zenoss redid the netapp zenpack.  almost as good as DFM.

[29-Sep-2011 14:43:04] <rmatte> cake php

[29-Sep-2011 14:43:10] <rmatte> oops

[29-Sep-2011 14:43:12] <rmatte> wrong window

[29-Sep-2011 14:43:28] <rmatte> hate that when I think I have focus on one window and I don't

[29-Sep-2011 14:43:45] <tvincent> rmatte: at least it wasn't "Root is ....."

[29-Sep-2011 14:43:56] <Guest97443>

[29-Sep-2011 14:43:56] <rmatte> haha

[29-Sep-2011 14:44:26] <Guest97443> any ideas abt netflow or sflow tools to intergrate with zenoss??

[29-Sep-2011 14:44:47] <tvincent> what do you want to do exactly? Have you looked at how other systems integrate with them?

[29-Sep-2011 14:44:51] <dhopp> rmatte: how long would you say it took you to write the daemon to integrate with Zenoss and OTRS (aside from the massive changes you probably did to OTRS)?

[29-Sep-2011 14:45:22] <rmatte> dhopp: a little over a week

[29-Sep-2011 14:46:41] <Guest97443> in simple words if i find any throughput is hiking up i will use netflow to check traffic flowing through... have u guys tried any netflow tools?? any recommendations?

[29-Sep-2011 14:47:00] <rmatte> Guest97443: this is about as good as it gets, and there really isn't much integration: http://www.plixer.com/products/netflow-sflow/scrutinizer-zenoss.php

[29-Sep-2011 14:47:18] <Guest97443> ya i used tht

[29-Sep-2011 14:47:25] <Guest97443> any other choice?

[29-Sep-2011 14:47:36] <Guest97443> i like it

[29-Sep-2011 14:47:41] <rmatte> well...

[29-Sep-2011 14:47:48] <rmatte> you're only using it for triage right?

[29-Sep-2011 14:47:54] <rmatte> you're not expecting to use netflow for alerting?

[29-Sep-2011 14:48:03] <rmatte> because you can alert using regular SNMP just as well

[29-Sep-2011 14:48:41] <Guest97443> one fuzzy abt it is i want constantly display graph in the dashboard of zenoss

[29-Sep-2011 14:49:10] <dhopp> doh! I can't seem to find this script I was working on it doesn't seem to be checked into svn :-(

[29-Sep-2011 14:49:15] <Guest97443> so i was doing it as displaying other graphs by mapping to it the link... but this one runs on flash

[29-Sep-2011 14:50:22] <Guest97443> but scrutinizer is fancy enough... i liked it.. but i m trying to see any other alternatives

[29-Sep-2011 15:19:12] <jmp242> Guest97443: We're going into some phase 2 testing with nTop

[29-Sep-2011 15:19:25] <jmp242> but we aren't really looking at using it in Zenoss so . . .

[29-Sep-2011 15:19:40] <jmp242> though if you used a site portlet, it probably doesn't matter?

[29-Sep-2011 15:22:20] <Guest97443> jmp242: oh ya i came across this tool during the search.. but didnt do the testing... how is it??

[29-Sep-2011 15:26:49] <jmp242> Well, it's good and bad

[29-Sep-2011 15:26:55] <jmp242> it's good in that it's free

[29-Sep-2011 15:26:59] <jmp242> or very cheap

[29-Sep-2011 15:27:13] <jmp242> bad in that I need to create a watcher to restart it when the web gui segfaults

[29-Sep-2011 15:27:16] <jmp242> but we use Puppet

[29-Sep-2011 15:27:20] <jmp242> so shouldn't be any big deal

[29-Sep-2011 15:27:34] <jmp242> It works well for what it does

[29-Sep-2011 15:27:40] <jmp242> and the nProbe works well enough also

[29-Sep-2011 15:34:43] <Guest97443> ohh ok cool... thanks for the info

[29-Sep-2011 15:34:52] <Guest97443> i will try tht

[29-Sep-2011 15:36:12] <dhopp> is there a way in zenbackup to tell it to not backup perf?

[29-Sep-2011 15:36:41] <rocket> dhopp: zenbackup --help

[29-Sep-2011 15:36:47] <rocket> dhopp: but yes there is

[29-Sep-2011 15:37:22] <dhopp> why use —help when I have #zenoss :-P

[29-Sep-2011 15:37:33] <Hackman238> dhopp: Nein!

[29-Sep-2011 15:37:39] <dhopp> ah —no-perfdata

[29-Sep-2011 15:37:44] <Hackman238> *taps on nose*

[29-Sep-2011 15:37:46] <Hackman238> LOL

[29-Sep-2011 15:38:05] <dhopp> Hackman238:  I probably use google before I use man pages..heh

[29-Sep-2011 15:38:10] * rocket teaches man to fish ...

[29-Sep-2011 15:38:14] <Hackman238> (like when my cat tried to drink from my cup*

[29-Sep-2011 15:38:26] <Hackman238> dhopp: Im only kidding

[29-Sep-2011 15:38:42] <dhopp> Hackman238:  so am I…I was just being lazy

[29-Sep-2011 15:39:07] <Hackman238> dhopp: I know how you mean

[29-Sep-2011 15:39:13] <Hackman238> dhopp: feels like friday

[29-Sep-2011 15:39:21] <dhopp> Hackman238: it's not?

[29-Sep-2011 15:39:22] <Hackman238> Someone change the date to friday

[29-Sep-2011 15:39:22] <dhopp> shit

[29-Sep-2011 15:39:30] <jmp242> Strawberry Icecream

[29-Sep-2011 15:39:38] <dhopp> is ding a zeopack every 4 hours too much?

[29-Sep-2011 15:39:39] <Hackman238> jmp242: ?

[29-Sep-2011 15:39:40] <jmp242> makes the afternoon better

[29-Sep-2011 15:39:42] <dhopp> *doing

[29-Sep-2011 15:39:53] <jmp242> Made some in my Blendtec

[29-Sep-2011 15:39:54] <Hackman238> dhopp: If you do it often each pack will take less time

[29-Sep-2011 15:39:54] <jmp242> mmmmmm

[29-Sep-2011 15:39:56] <jmp242>

[29-Sep-2011 15:40:08] <Hackman238> jmp242: nice. Did you see if the iphone4 blends?

[29-Sep-2011 15:40:18] <jmp242> no

[29-Sep-2011 15:40:29] <jmp242> I think my co workers would kill me if I blended their iPhone

[29-Sep-2011 15:40:38] <dhopp> Hackman238:  I was thinking of doing zeopack every 4 hours and then saving a couple of days worth of the .old files

[29-Sep-2011 15:40:40] <Hackman238> jmp242: Have you seen the 'does it blend' blendtec videos?

[29-Sep-2011 15:40:45] <jmp242> yes

[29-Sep-2011 15:40:47] <dhopp> of course that depends on how big the .old file is

[29-Sep-2011 15:40:51] <jmp242> It's one of the reasons I bought it

[29-Sep-2011 15:40:57] <Hackman238> dhopp: Not a bad idea long as they arent huge

[29-Sep-2011 15:41:09] <Hackman238> jmp242: Great advertising idea

[29-Sep-2011 15:41:14] <jmp242> Yes indeed

[29-Sep-2011 15:41:26] <jmp242> And the thing lives up to the videos for food anyway

[29-Sep-2011 15:41:30] <jmp242> whole carrots etc

[29-Sep-2011 15:42:07] <Hackman238> jmp242: I tried on an assortment of old sprint flip phones and nokia trac phones

[29-Sep-2011 15:42:22] <jmp242> yea

[29-Sep-2011 15:42:51] <Hackman238> jmp242: Ate them like nothing. Sorta scarey.

[29-Sep-2011 15:42:52] <jmp242> did you have a second pitcher

[29-Sep-2011 15:42:56] <jmp242> or how did you clean it?

[29-Sep-2011 15:43:21] <Hackman238> jmp242: I jsut got a new pitcher. the contaminated one is for fun and chemistry purposes only

[29-Sep-2011 15:43:31] <jmp242> lol

[29-Sep-2011 15:43:39] <jmp242> I am considering it for making wood chips lol

[29-Sep-2011 15:43:46] <Hackman238> jmp242: I should note I did take the Li and LiPo batteries out before doing it.

[29-Sep-2011 15:43:50] <Hackman238> didnt really want a fire

[29-Sep-2011 15:43:52] <jmp242> oh yea

[29-Sep-2011 15:43:55] <jmp242> that's a good idea

[29-Sep-2011 15:44:00] <jmp242> did you see the time warp video

[29-Sep-2011 15:44:01] <Hackman238> jmp242: For like a smoker?

[29-Sep-2011 15:44:03] <jmp242> with the bic lighters

[29-Sep-2011 15:44:04] <jmp242> ?

[29-Sep-2011 15:44:22] <Hackman238> jmp242: No, sounds awesome

[29-Sep-2011 15:44:40] <jmp242> the thing kept going after the flame explosion and fire - no problem just singed a bit

[29-Sep-2011 15:45:03] <Hackman238> jmp242: Speaking of wood chips and smoked items, Laphoaig Scotch is a smoked scotch you must try

[29-Sep-2011 15:45:09] <jmp242> The blendtec is one of my favorite kitchen gadgets

[29-Sep-2011 15:45:19] <Hackman238> jmp242: Nice!

[29-Sep-2011 15:45:19] <jmp242> the other is my Bosch Universal Mixer

[29-Sep-2011 15:45:59] <jmp242> Actually, if I had gotten that first, I might not have gotten the blendtec and instead gone with the blender add on for that

[29-Sep-2011 15:46:17] <Hackman238> jmp242: I've been known to use my drill with a wide chuck to mix things. LOL

[29-Sep-2011 15:47:13] <jmp242> Lol

[29-Sep-2011 15:47:31] <jmp242> The mixer is pretty awesome, can do like 17lbs of fresh ground whole wheat dough at once

[29-Sep-2011 15:47:43] <jmp242> 22 cups of flour lol

[29-Sep-2011 15:47:57] <Hackman238> jmp242: Damn. I dont even know what I'd do with that much dough.

[29-Sep-2011 15:48:16] <jmp242> make 9 loaves of bread

[29-Sep-2011 15:48:22] <jmp242> and freeze them or give them away

[29-Sep-2011 15:48:22] <Hackman238> jmp242: put it in a bag and drop off a 4 story building on to a packed cars hood manybe

[29-Sep-2011 15:48:39] <Hackman238> jmp242: LOL I know not how to make bread.

[29-Sep-2011 15:48:50] <Hackman238> *parked

[29-Sep-2011 15:48:53] <jmp242> Well, I usually do about 6 loaves, though may just go to the 3 2lb loaves which are grocery store sized

[29-Sep-2011 15:49:04] <Hackman238> jmp242: Ah.

[29-Sep-2011 15:49:23] <jmp242> Bread isn't that hard for me - baking is like computers - you follow the recipe and you get the result

[29-Sep-2011 15:49:31] <Hackman238> jmp242: Thats a lot of bread. I dont think I eat a loaf of bread in a month

[29-Sep-2011 15:49:33] <jmp242> other cooking is harder as it's much less specific

[29-Sep-2011 15:49:46] <jmp242> Oh, well I give a bunch to co workers and friends

[29-Sep-2011 15:49:54] <Hackman238> jmp242: Ah

[29-Sep-2011 15:50:05] <jmp242> it's fun to do, healthier and tastes like 100 times better than store bought whole wheat (which I won't eat)

[29-Sep-2011 15:50:19] <jmp242> I do grind the wheat berries in my blendtec though

[29-Sep-2011 15:50:24] <jmp242> to make the flour

[29-Sep-2011 15:50:30] <jmp242> until I save up for the Nutrimill

[29-Sep-2011 15:50:44] <jmp242> cause I can only do ~4 cups of flour at a time in the blendtec even with the wildside jar

[29-Sep-2011 15:51:16] <dhopp> so if I use zenbackup to backup everything but perf and mysql…and then later restore it..there shouldn't be any problems with mysql having events that didn't exist when the backup was taken?

[29-Sep-2011 15:51:35] <Hackman238> jmp242: Make your own flour?

[29-Sep-2011 15:51:43] <Hackman238> jmp242: Interesting.

[29-Sep-2011 15:52:14] <dhopp> Hackman238:  how are you guys backing up your mysql databases?  I can't imagine you are doing mysqldump

[29-Sep-2011 15:52:35] <dhopp> I'm thinking stop mysql, take lvm snapshot, start mysql, backup snapshot

[29-Sep-2011 15:52:39] <Hackman238> dhopp: We have a replication partner setup that we dump from freqeuntly

[29-Sep-2011 15:52:44] <dhopp> (on a second host..not the primary)

[29-Sep-2011 15:53:02] <dhopp> you do a mysqldump frequently?

[29-Sep-2011 15:53:09] <Hackman238> dhopp: Yes

[29-Sep-2011 15:53:17] <dhopp> don't that take forever on the size of database you guys probably have?

[29-Sep-2011 15:53:21] <dhopp> *doesn't

[29-Sep-2011 15:53:24] <dhopp> I talk good

[29-Sep-2011 15:53:25] <Hackman238> dhopp: yep.

[29-Sep-2011 15:53:55] <Hackman238> dhopp: don't sweat it. I probably typo in here 2x more than anyone else

[29-Sep-2011 15:54:15] <dhopp> Hackman238:  of course you talk like 4x more than anyone :-P

[29-Sep-2011 15:54:38] <Hackman238> LOL

[29-Sep-2011 15:55:13] <jmp242> oh making your own flour is the only way to go for bread

[29-Sep-2011 15:55:22] <jmp242> well, if you want whole wheat

[29-Sep-2011 15:55:27] <jmp242> plus it's fun lol

[29-Sep-2011 15:55:58] <jmp242> how often are you all backing up Zenoss? We do it daily ourselves

[29-Sep-2011 15:56:10] <dhopp> Hackman238:  I remember doing mysqldump on like 30GB database and it took a long time…wouldn't that stop replication while it's running?

[29-Sep-2011 15:56:30] <dhopp> jmp242:  I will do it more often to disk and then daily to tape

[29-Sep-2011 15:56:42] <Hackman238> dhopp: It doesnt seem to cause a problem

[29-Sep-2011 15:56:59] <Hackman238> jmp242: Might be a fun thing to try

[29-Sep-2011 15:57:20] <dhopp> Hackman238:  I'll need to investigate..maybe it's only on MyISAM that it will lock the table

[29-Sep-2011 15:57:54] <jmp242> Speaking of MySQL - is anyone thinking about the current kerfluffle re Oracle and Postgres or something?

[29-Sep-2011 15:57:59] <Hackman238> jmp242: So if you open your mail box and theres a tube the diameter of a loaf of bread, its definately bread even if its hard as concrete or black as cal from over cooking

[29-Sep-2011 15:58:22] <Hackman238> jmp242: I tihnk Zenoss is considering a move to postgres

[29-Sep-2011 15:58:36] <Hackman238> s/cal/coal/

[29-Sep-2011 16:01:33] <dhopp> heh…When using http://dev.mysql.com/doc/refman/5.5/en/mysqldump.html, you should stop replication on the slave before starting the dump process to ensure that the dump contains a consistent set of data:

[29-Sep-2011 16:03:38] <Hackman238> dhopp: interesting. We'll have to add that to the backup script

[29-Sep-2011 16:04:18] <dhopp> Hackman238:  that's according to the official docs...http://dev.mysql.com/doc/refman/5.5/en/replication-solutions-backups-mysqldump.html

[29-Sep-2011 16:04:28] <dhopp> not sure if it's actually necessary..but a best practice

[29-Sep-2011 16:07:16] <Hackman238> dhopp: Gotcha

[29-Sep-2011 16:14:08] <dhopp> I might look at using the free version of Zmanda Recovery Manager…I've used it in the past for smaller databases...

[29-Sep-2011 16:16:16] <tvincent> i could see zenoss using other db's as an option. But replacing mysql would be nuts.

[29-Sep-2011 16:17:32] <rmatte> it's too bad that Zenoss wasn't built with an actual database layer

[29-Sep-2011 16:17:33] <dhopp> I don't think there is anything wrong with using MySQL…just wish MySQL had a better backup option…I keep hearing about a backup api that is supposed to be in MySQL but that seems to never make it into a release

[29-Sep-2011 16:17:37] <jb> why would it be nuts?

[29-Sep-2011 16:17:41] <rmatte> would have made it a lot easier to support different database types

[29-Sep-2011 16:17:54] <jb> mysql is gone in 4.2, I believe.

[29-Sep-2011 16:18:09] <dhopp> what did they move to?

[29-Sep-2011 16:18:11] <rmatte> what are they replacing it with?

[29-Sep-2011 16:18:13] <jb> postgres

[29-Sep-2011 16:18:18] <rmatte> ah

[29-Sep-2011 16:18:24] <Hackman238> mysql licensing is a problem, probably going to psotgres

[29-Sep-2011 16:18:24] <rmatte> smart move, much faster

[29-Sep-2011 16:18:29] <Hackman238> what jb said

[29-Sep-2011 16:18:32] <Hackman238> LOL Sorry

[29-Sep-2011 16:18:34] <jb> it's really a licensing issue

[29-Sep-2011 16:18:36] <jb> what hackman said

[29-Sep-2011 16:18:37] <jb> hah

[29-Sep-2011 16:18:46] <dhopp> what's the licensing issue?

[29-Sep-2011 16:18:50] <jb> oracle

[29-Sep-2011 16:18:53] <Hackman238> dhopp: mysql with oracle

[29-Sep-2011 16:19:19] <Hackman238> dhopp: they terrorize Zenoss with costly per instance fees

[29-Sep-2011 16:19:27] <dhopp> I guess I don't follow…

[29-Sep-2011 16:19:31] <dhopp> oh..for support you mean

[29-Sep-2011 16:20:04] <Hackman238> dhopp: Zenoss cant use mysql in a commercial capacity in the product unless its licensed

[29-Sep-2011 16:20:46] <dhopp> isn't mysql GPL though?

[29-Sep-2011 16:21:22] <Hackman238> dhopp: I dont think so anymore

[29-Sep-2011 16:21:28] <dhopp> ah

[29-Sep-2011 16:21:37] <jb> btw, at least in the ent SD versions, it's mysql enterprise

[29-Sep-2011 16:21:50] <Hackman238> Yep

[29-Sep-2011 16:22:03] <Hackman238> its so boss the differnce is unnoticable

[29-Sep-2011 16:22:11] <Hackman238> LOL

[29-Sep-2011 16:22:13] <jb> haha

[29-Sep-2011 16:22:15] <dhopp> "For Commercial OEMs, ISVs and VARs:

[29-Sep-2011 16:22:15] <dhopp>           •          For OEMs, ISVs, and VARs who distribute MySQL with their products, and do not license and distribute their source code under the GPL, MySQL provides a flexible OEM Commercial License.

[29-Sep-2011 16:22:15] <dhopp> "

[29-Sep-2011 16:22:36] <Hackman238> "Flexible" means "bend over"

[29-Sep-2011 16:22:40] <Hackman238> XD

[29-Sep-2011 16:22:44] <dhopp> "If you are developing and distributing open source applications under the GPL License, then you are free to use MySQL under the GPL License."

[29-Sep-2011 16:22:59] <dhopp> Maybe Zenoss should just release everything under the GPL then :-P

[29-Sep-2011 16:23:05] <Hackman238> dhopp: Ent is mixed license

[29-Sep-2011 16:23:32] <dhopp> Hackman238:  I know..but change it to GPL and no more licensing fees :-)

[29-Sep-2011 16:23:49] <Hackman238> dhopp: true, but I dont see that happening

[29-Sep-2011 16:24:06] <dhopp> Hackman238:  I know..I'm kidding

[29-Sep-2011 16:24:08] <dhopp> but one can dream

[29-Sep-2011 16:24:29] <Hackman238> I do like the use of the word "flexible" the the Oracle statement.

[29-Sep-2011 16:25:08] <Hackman238> Because the costs are so high its madness. Really is an inside joke for "bend over suckers!"

[29-Sep-2011 16:25:37] <Hackman238> I can only imagine the lawyer smirking as he thought up that line

[29-Sep-2011 16:26:55] <dhopp> flexible is a great word in legal documents..but it pretty much can mean anything

[29-Sep-2011 16:27:11] <dhopp> "wait you changed the price for support?"…"well our pricing is 'flexible'"

[29-Sep-2011 16:28:04] <dhopp> so a prospective client reads that and says "I can talk them down"…when in reality it means "we will charge whatever we want"..heh

[29-Sep-2011 16:29:32] <Hackman238> dhopp: true true

[29-Sep-2011 16:30:08] <jmp242> so what, does the relstorage work in Postgres/

[29-Sep-2011 16:30:09] <jmp242> ?

[29-Sep-2011 16:31:26] <Hackman238> jmp242: I dont know.

[29-Sep-2011 16:31:32] <Hackman238> jmp242: Its doable

[29-Sep-2011 16:31:34] <dhopp> jmp242:  http://pypi.python.org/pypi/RelStorage

[29-Sep-2011 16:31:37] <dhopp> looks like it should

[29-Sep-2011 16:32:03] <jb> anybody here good with sed/text manipulation?

[29-Sep-2011 16:32:16] <dhopp> jb: what are you trying to do?

[29-Sep-2011 16:33:01] <jb> i have a continuous stream of text between "--- start ---" and "+++ end +++"

[29-Sep-2011 16:33:21] <jb> i can extract that with something like sed -n '/---/,/+++/p'

[29-Sep-2011 16:33:22] <dhopp> ok....

[29-Sep-2011 16:33:42] <jb> the problem is, is there there are multiple sections each delimited with the --- and +++'s in the file.

[29-Sep-2011 16:33:49] <jb> i ONLY need to get the very first one

[29-Sep-2011 16:33:57] <jb> which can be any number of lines..

[29-Sep-2011 16:35:55] <mahimahi> Hi. I'm having problems with monitoring template that return multiple values not displaying in graph zenoss v3.2.  Anyone familiar with this?

[29-Sep-2011 16:37:56] <dhopp> jb: I've done something similar before..trying to remember how

[29-Sep-2011 16:38:10] <rmatte> mahimahi: you can't have an OID returning multiple values in a template, it doesn't work

[29-Sep-2011 16:38:19] <rmatte> mahimahi: it has to be an exact OID, not a tree

[29-Sep-2011 16:38:39] <jbk> jb: why not awk 'BEGIN { do_print = 0} /^--- start/ { do_print=1 } /^+++ end/ { exit; } { if (do_print) { print } }'

[29-Sep-2011 16:38:41] <mahimahi> this is not a SNMP, its a custom like check_http

[29-Sep-2011 16:38:54] <rmatte> mahimahi: oh, did you install Zenoss 3.2?

[29-Sep-2011 16:39:11] <mahimahi> yes

[29-Sep-2011 16:39:18] <rmatte> did you patch it with these patches...?

[29-Sep-2011 16:39:22] <rmatte> zenpatch 26334

[29-Sep-2011 16:39:27] <rmatte> zenpatch 26335

[29-Sep-2011 16:39:31] <rmatte> zenpatch 26337

[29-Sep-2011 16:39:39] <mahimahi> used to work in 3.1, upgrade to 3.2 then lost my instance

[29-Sep-2011 16:39:41] <rmatte> you need to do that as the Zenoss user then restart Zenoss

[29-Sep-2011 16:39:47] <mahimahi> nope, that's probably it

[29-Sep-2011 16:39:48] <rmatte> right, because you need those patches

[29-Sep-2011 16:39:52] <rmatte>

[29-Sep-2011 16:39:53] <mahimahi> thanks

[29-Sep-2011 16:39:56] <rmatte> no prob

[29-Sep-2011 16:55:53] <Hackman238> Have to run

[29-Sep-2011 16:55:58] <rmatte> ah, later

[29-Sep-2011 16:55:59] <Hackman238> Later all

[29-Sep-2011 17:09:59] <Guest97443> by default, in zenoss dashboard the layout can divided to three columns but in my case i want to add more graphs in my dashboard like small blocks of required graphs of the devices... any idea to tweak it... how u guys set it up ??

[29-Sep-2011 17:10:03] <Guest97443> any ideas?

[29-Sep-2011 17:10:32] <rmatte> I don't really understand what you mean

[29-Sep-2011 17:10:49] <rmatte> like you want to add specific device graphs to it?

[29-Sep-2011 17:11:17] <rmatte> use the site window portlet, then right click on the graph you want to add and copy the location, then use that location in the site window portlet

[29-Sep-2011 17:11:21] <rmatte> repeat for each graph you wanty

[29-Sep-2011 17:11:23] <rmatte> want*

[29-Sep-2011 17:11:33] <Guest97443> u know tht we can display perf graph in the dashboard...

[29-Sep-2011 17:11:34] <Guest97443> yaaa

[29-Sep-2011 17:11:36] <Guest97443> exactly

[29-Sep-2011 17:11:40] <Guest97443> thts wht i m doing

[29-Sep-2011 17:11:51] <rmatte> that's the only way to do it

[29-Sep-2011 17:12:01] <Guest97443> but i want to display more of graphs but layout itself can break to three

[29-Sep-2011 17:12:24] <Guest97443> so there is limitation of adding more graphs to display

[29-Sep-2011 17:12:30] <rmatte> you can't break it any further if that's what you mean

[29-Sep-2011 17:12:32] <rmatte> 3 is the limit

[29-Sep-2011 17:12:38] <Guest97443> yaa

[29-Sep-2011 17:12:43] <rmatte> there is no limitation

[29-Sep-2011 17:12:49] <rmatte> just add them on new rows

[29-Sep-2011 17:13:10] <rmatte> you're limited to 3 columns but you can have as many rows as you want

[29-Sep-2011 17:13:45] <Guest97443> ahhhh i hav to fix my big screen verticals then

[29-Sep-2011 17:14:26] <Guest97443> how u do in ur setup for dashboard?? u display graphs in dashboard

[29-Sep-2011 17:14:28] <Guest97443> ?

[29-Sep-2011 17:14:37] <rmatte> yeh, if you have some huge resolution it won't work very well

[29-Sep-2011 17:14:58] <rmatte> you could always just code your own custom website to display the graphs

[29-Sep-2011 17:16:47] <Guest97443> hmm initially i was thinking of there is some option to remove tht limitation and custom the graph size to smaller and display many number.... instead going through custom website creation option... ahhh i think i hav to use ur idea of creating custom website :S

[29-Sep-2011 17:16:58] <jbk> is there an easy way to grab the rrd data from zenoss via http?

[29-Sep-2011 17:17:35] <rmatte> jbk: a REST call can be used for that

[29-Sep-2011 17:17:46] <jbk> ahh perfect

[29-Sep-2011 17:18:02] * jbk would like to rig something up with sparklines

[29-Sep-2011 17:18:13] <rmatte> jbk: example... http://USERNAME:PASSWORD@MY_ZENOSS_HOST:8080/zport/dmd/Devices/Server/Linux/devices/angel/getRRDValue?dsname=laLoadInt5_laLoadInt5

[29-Sep-2011 17:18:27] <rmatte> if that's what you mean

[29-Sep-2011 17:18:33] <rmatte> it'll return the most recent RRD value

[29-Sep-2011 17:18:52] <jbk> what about a range of datapoints?

[29-Sep-2011 17:19:30] <rmatte> I really don't think there's any built in method to do that

[29-Sep-2011 17:19:51] <jbk> kinda like rrdtool xport

[29-Sep-2011 17:20:11] <rmatte> there's no way to easily do that via HTTP

[29-Sep-2011 17:20:20] <rmatte> you could write a python script to do it

[29-Sep-2011 17:21:40] <rmatte> you'd do: d.getRRDValues('laLoadInt5_laLoadInt5', start=startDateInEpoch, end=endDateInEpoch)

[29-Sep-2011 17:21:52] <rmatte> and you'd have to adjust the start and end dates for each period you want to grab for

[29-Sep-2011 17:22:04] <rmatte> and cycle through it

[29-Sep-2011 17:22:34] <rmatte> or just actually use rrdtool exports

[29-Sep-2011 17:33:36] <jbk> but no easy way to (for example) put the xml output of rrdtool xport into a web page?

[29-Sep-2011 17:35:05] <jbk> since i could grab the values via js

[29-Sep-2011 17:35:22] <rmatte> no, there's no built in way to do that

[29-Sep-2011 17:35:27] <rmatte> you'd have to code something custom

[29-Sep-2011 17:35:44] <dhopp> rmatte: didn't Hackman238 have something that did that..but only worked in 2.x?

[29-Sep-2011 17:35:59] <rmatte> no idea, if he does I've never heard of it

[29-Sep-2011 17:36:39] <dhopp> I thought he sent it to you…where he added calendar option when looking at a graph

[29-Sep-2011 17:36:45] <dhopp> maybe not

[29-Sep-2011 17:36:52] <rmatte> dhopp: that's not what jbk is asking for

[29-Sep-2011 17:36:57] <dhopp> oh

[29-Sep-2011 17:37:04] <rmatte> he wants to be able to call a URL and get a dump of every data point within a specified range

[29-Sep-2011 17:37:15] <rmatte> so literally 1,67,89,45,67,87,34,etc...

[29-Sep-2011 17:37:17] <jbk> i want to get the datapoints into a webpage

[29-Sep-2011 17:37:29] <dhopp> ohhh..sorry I caught the tail end of the conversation

[29-Sep-2011 17:37:40] <rmatte> all good

[29-Sep-2011 17:37:48] <rmatte> and Hackman hasn't sent me that pack yet either

[29-Sep-2011 17:37:53] <rmatte> lol

[29-Sep-2011 17:37:55] <dhopp> rmatte:  bastard

[29-Sep-2011 17:37:57] <jbk> dhopp: i want to use it with this: http://omnipotent.net/jquery.sparkline/

[29-Sep-2011 17:38:50] <rmatte> hmmm, those would be badass in Zenoss

[29-Sep-2011 17:39:05] <rmatte> when viewing interfaces or whatever, graph out the utilization on each row

[29-Sep-2011 17:39:13] <jbk> (i've read a few of tufte's books and went to his seminar :P)

[29-Sep-2011 17:40:02] <rmatte> cool

[29-Sep-2011 17:40:15] <SDuensin> Would it be a Bad Idea(tm) to have a modeler update the "Device Name" field?

[29-Sep-2011 17:40:34] <rmatte> SDuensin: could be, why would you want to?

[29-Sep-2011 17:40:47] <jb> ugh

[29-Sep-2011 17:41:00] <jb> the "test command" funtion in a zencommand data source's dialog box

[29-Sep-2011 17:41:03] <SDuensin> So I have actual device names instead of discovered IP addresses.

[29-Sep-2011 17:41:08] <jb> does not actually test the command against the remote device via SSH

[29-Sep-2011 17:41:14] <jb> it tries to test the device against the zenoss box it's self.

[29-Sep-2011 17:41:24] <jb> test the command rather

[29-Sep-2011 17:41:27] <dhopp> SDuensin:  http://www.shanewilliamscott.com/ look at the Snmp Name Modeler

[29-Sep-2011 17:41:39] <dhopp> I haven't tried it personally

[29-Sep-2011 17:41:46] <SDuensin> dhopp, will do, but I'm kind of rolling my own here.

[29-Sep-2011 17:42:11] <dhopp> SDuensin:  you can look at his code and might give you some ideas

[29-Sep-2011 17:42:20] <rmatte> jb: correct, to test you should be doing: zencommand run -v10 -d <devicename>

[29-Sep-2011 17:42:23] <SDuensin> dhopp, code is Good.  :-)

[29-Sep-2011 17:42:52] <dhopp> SDuensin:  personally I'm a little miffed that when you add a device and you set an IP address AND name that if it can't resolve via dns it defaults to using the IP address…I GAVE YOU A NAME..USE IT!

[29-Sep-2011 17:43:26] <rmatte> dhopp: meh, I always just change the name after adding the devices

[29-Sep-2011 17:43:56] <dhopp> rmatte:  sure you can do that..but why can it just do that when you add the device..why should it be a 2 step process?

[29-Sep-2011 17:44:17] <rmatte> no idea

[29-Sep-2011 17:44:26] <dhopp> rmatte:  my point exactly

[29-Sep-2011 17:44:33] <rmatte> there are many things in Zenoss that make little sense lol

[29-Sep-2011 17:44:42] <rmatte> but you learn to live with them since most aren't very major

[29-Sep-2011 17:44:58] <rmatte> and overall it's still miles ahead of most other products

[29-Sep-2011 17:45:52] <dhopp> I mean it shouldn't be that hard to do "if Title is null then do dns lookup if dns lookup fails use whatever is in "Name or IP" box"

[29-Sep-2011 17:46:26] <rmatte> dhopp: well actually, you might be confused...

[29-Sep-2011 17:46:45] <rmatte> that input box where it says "Device Name" on the device status page isn't actually the device name

[29-Sep-2011 17:46:45] <jb> hrm, I bound a template to /Server/Linux/DNS, but devices in /Server/Linux/DNS/External are not inheriting it.

[29-Sep-2011 17:46:48] <rmatte> it's the device title

[29-Sep-2011 17:46:58] <rmatte> to change the device name, you have to use the menu option in the bottom left

[29-Sep-2011 17:47:04] <jb> must be something silly.

[29-Sep-2011 17:47:35] <rmatte> jb: because /Server/Linux/DNS/External probably already has a custom template binding

[29-Sep-2011 17:47:41] <dhopp> rmatte: I know..but is special about "title" then?  I don't know why you would set title to anything other than the device name

[29-Sep-2011 17:47:47] <dhopp> *what is

[29-Sep-2011 17:47:51] <rmatte> go in zProperties for /Server/Linux/DNS/External, scroll all the way to the bottom, you'll see a dropdown

[29-Sep-2011 17:47:59] <rmatte> check and see if zDeviceTemplates is in that dropdown

[29-Sep-2011 17:48:04] <rmatte> if it is then it has a locally set one

[29-Sep-2011 17:48:13] <rmatte> you need to delete the local setting to inherit from the parent

[29-Sep-2011 17:48:27] <jb> hrm, which zprop?

[29-Sep-2011 17:48:37] <rmatte> zDeviceTemplate

[29-Sep-2011 17:48:53] <jb> yeah it does have one

[29-Sep-2011 17:49:10] <jb> gotcha

[29-Sep-2011 17:49:11] <jb> thanks

[29-Sep-2011 17:49:18] <rmatte> dhopp: the way it works is this... you may have the device name as the full hostname, blah.blah.com, but want the device name to actually be NEWYORK01

[29-Sep-2011 17:49:33] <rmatte> dhopp: so you set the title to NEWYORK01 and that's what will show up when you look in lists in the UI

[29-Sep-2011 17:49:39] <rmatte> but it's still technically called by the hostname

[29-Sep-2011 17:50:03] <rmatte> title overrides device name when viewing the device name in any UI screen

[29-Sep-2011 17:50:15] <rmatte> jb: no prob

[29-Sep-2011 17:50:20] <dhopp> rmatte:  ok I get that…but why not have everything tied to the title (i.e. the directory in perf/Devices) ?

[29-Sep-2011 17:50:39] <rmatte> dhopp: because the device name is the absolute root of everything

[29-Sep-2011 17:50:42] <rmatte> it always has been

[29-Sep-2011 17:50:51] <rmatte> title is just something they added as an option

[29-Sep-2011 17:50:54] <rmatte> I personally never use it

[29-Sep-2011 17:51:10] <rmatte> and I've been saying they should change it to say Device Title: instead of Device Name:

[29-Sep-2011 17:51:34] <dhopp> rmatte: I get that the device is the root of everything

[29-Sep-2011 17:51:53] <dhopp> rmatte:  in 3.x when you add a device all it says is 'Title' so it's sort of misleading

[29-Sep-2011 17:52:13] <rmatte> I agree, it's extremely misleading, but that's how they decided to code it

[29-Sep-2011 17:52:22] <rmatte> yet another UI usability issue that will probably take over a year to address

[29-Sep-2011 17:52:36] <rmatte> it took 5 months to agree to have them put the word "Copy" back on the Override button

[29-Sep-2011 17:52:41] <dhopp> rmatte:  to me it just seems to have made sense to do something like 'if dns failure and Title is set use Title'..but it makes no sense to have a device name of '10.10.10.12'

[29-Sep-2011 17:53:11] <rmatte> just don't use title at all is my point

[29-Sep-2011 17:53:20] <rmatte> title is only for very special circumstances

[29-Sep-2011 17:53:23] <rmatte> just stick to using the device name

[29-Sep-2011 17:53:43] <rmatte> leave title blank

[29-Sep-2011 17:53:56] <dhopp> rmatte:  and that's fine..but in an environment where the device is not in DNS then you have a 2 step process to add the device..either add the device name and let it fail and then change the manage IP..or add by ip and change the device name

[29-Sep-2011 17:53:57] <rmatte> and as for the auto-dns resolution, I believe that is actually an option, you just probably have it disabled

[29-Sep-2011 17:54:00] <rmatte> hold on, I'll find the setting

[29-Sep-2011 17:54:51] <dhopp> rmatte:  My dns names will work IF the device is in DNS (or the host file) but for reasons unknown to me almost none of our devices are in DNS

[29-Sep-2011 17:55:02] <dhopp> so I can resolve mycooldevice.domain.com

[29-Sep-2011 17:55:04] <dhopp> err

[29-Sep-2011 17:55:04] <dhopp> can't

[29-Sep-2011 17:55:11] <rmatte> ok, well as I've explained, title is not device name

[29-Sep-2011 17:55:13] <rmatte> so that's why

[29-Sep-2011 17:55:21] <rmatte> leave title blank and then rename the device by hand after adding it

[29-Sep-2011 17:55:31] <rmatte> using the bottom left hand menu

[29-Sep-2011 17:55:36] <rmatte> "Rename Device"

[29-Sep-2011 17:55:56] <dhopp> rmatte:  and I get what you are saying..I just think it's stupid...

[29-Sep-2011 17:56:04] <rmatte> no argument there

[29-Sep-2011 17:56:07] <rmatte> open a ticket about it

[29-Sep-2011 17:56:11] <rmatte> dev.zenoss.org/trac

[29-Sep-2011 17:56:14] <rmatte> zenoss/zenoss to login

[29-Sep-2011 17:56:17] <dhopp> rmatte:  and I know when people start to use this they are going to give me "WTF?  that's stupid"

[29-Sep-2011 17:56:27] <rmatte> yup

[29-Sep-2011 17:56:33] <dhopp> I know how to open tickets :-P

[29-Sep-2011 17:57:07] <rmatte> I wasn't making the assumption that you did or didn't, I just explain it by habbit in case you didn't

[29-Sep-2011 17:58:16] <rmatte> I've been answering questions in this channel for over 2 years, certain habbits grow on you

[29-Sep-2011 17:58:22] <rmatte> lol

[29-Sep-2011 17:58:39] <Guest97443> wow thts a long time

[29-Sep-2011 17:59:04] <dhopp> rmatte:  of course my response to them will be "Why don't we have our devices in DNS?  That's stupid"

[29-Sep-2011 17:59:05] <dhopp> hehe

[29-Sep-2011 17:59:07] <rmatte> and I've actually been in the channel for close to 3 years, but for the first year I was doing more asking than answering

[29-Sep-2011 17:59:17] <rmatte> dhopp: true

[29-Sep-2011 17:59:51] <Guest97443> rmatte: i m in tht phase

[29-Sep-2011 17:59:58] <Guest97443> ur 1st year

[29-Sep-2011 18:00:22] <rmatte> you don't become an expert overnight

[29-Sep-2011 18:00:30] <Guest97443> true

[29-Sep-2011 18:00:40] <dhopp> rmatte:  I was in this channel 3 or 4 years ago but was more just lurking…then I got back into using Zenoss and now trying to implement it monitoring the most devices I have ever needed to

[29-Sep-2011 18:01:02] <rmatte> dhopp: fun, 10,000 devices or something?

[29-Sep-2011 18:01:37] <dhopp> not that many

[29-Sep-2011 18:01:38] <dhopp> lol

[29-Sep-2011 18:01:46] <rmatte> how many?

[29-Sep-2011 18:01:48] <dhopp> ~1200

[29-Sep-2011 18:02:01] <dhopp> I think the most I've done in Zenoss before is about ~250

[29-Sep-2011 18:02:05] <rmatte> ah, that's about the number that I monitor, only mine are spread across 15 separate instances

[29-Sep-2011 18:02:20] <dhopp> yeah we are going to have 2 collectors and 1 hub machine

[29-Sep-2011 18:02:37] <rmatte> I have one server with 455 devices on it

[29-Sep-2011 18:03:06] <rmatte> 455 devices spread across 2 continents

[29-Sep-2011 18:03:08] <rmatte> http://dmon.org/graphics/zenoss/gen2.png

[29-Sep-2011 18:03:29] <dhopp> and like 20 countries

[29-Sep-2011 18:03:31] <dhopp> hehe

[29-Sep-2011 18:03:33] <rmatte> yup

[29-Sep-2011 18:03:54] <dhopp> these 1200 devices are in 2 datacenters...

[29-Sep-2011 18:04:40] <dhopp> and I'm trying to break the habit of "we need to know everything about everything"…uhm we do not need to monitor development machines at the same level as production

[29-Sep-2011 18:05:04] <rmatte> obviously not lol

[29-Sep-2011 18:05:25] <Guest97443> about tht google maping... I know tht we have get the key from google map to enable it... i always wonder is it safe?? does it speak with google often?? as i dont the internal zen to speak with outside environment...

[29-Sep-2011 18:05:51] <rmatte> Guest97443: why wouldn't it be safe?

[29-Sep-2011 18:06:27] <dhopp> Guest97443:  I don't use it since all of my devices are on a private network…Google wouldn't really be able to tell me where they are…and they are in 2 DCs so not that hard to find it..unless Google can tell me what Cage/Rack as well

[29-Sep-2011 18:06:28] <dhopp> heh

[29-Sep-2011 18:06:42] <dhopp> rmatte:  some people fear what Google is doing with that data

[29-Sep-2011 18:06:44] <rmatte> Guest97443: I'm pretty sure that it's technically your browsers talking to the google servers, not the Zenoss server itself

[29-Sep-2011 18:06:49] <Guest97443> Coz I dont want someone to use the port to get into box from outside as google... like a man in middle attack kinda

[29-Sep-2011 18:07:04] <rmatte> Guest97443: I highly doubt anything like that is possible

[29-Sep-2011 18:07:32] <dhopp> rmatte:  you got devices in Chicago...

[29-Sep-2011 18:07:35] <rmatte> Guest97443: you're just passing google a bunch of coordinates and they are spitting back a map, there's nothing that could possibly be used to get in to Zenoss

[29-Sep-2011 18:07:42] <dhopp> equinix datacenter by chance?

[29-Sep-2011 18:07:45] <rmatte> dhopp: our client has devices in chicago

[29-Sep-2011 18:07:53] <rmatte> nah, they are actually hosted at the client's site

[29-Sep-2011 18:08:01] <dhopp> ah

[29-Sep-2011 18:08:02] <dhopp> ok

[29-Sep-2011 18:08:04] <dhopp> I got to run

[29-Sep-2011 18:08:05] <dhopp> see ya

[29-Sep-2011 18:08:39] <rmatte> later

[29-Sep-2011 18:08:44] <rmatte> bah, too late lol

[29-Sep-2011 18:12:33] <Guest97443> hmm k then..

[29-Sep-2011 18:26:24] <tvincent> anyone else ever run into the problem where a machine's ip change

[29-Sep-2011 18:26:29] <tvincent> changes over night

[29-Sep-2011 18:26:43] <rmatte> can't say I have

[29-Sep-2011 18:26:44] <tvincent> and zenoss doesn't pickup that the devices ip has changed

[29-Sep-2011 18:26:51] <rmatte> oh

[29-Sep-2011 18:26:56] <rmatte> ummm, how can it and why should it?

[29-Sep-2011 18:26:58] <jb> hrmmm

[29-Sep-2011 18:27:04] <jb> 2011-09-29 17:24:31,854 DEBUG zen.zencommand: Queued event (total of 3) {'manager': 'fc-wrzenm01.corp.follett.com', 'eventKey': 'bindStats', 'device': 'dns1.follett.com', 'eventClass': '/Cmd/Fail', 'summary': '', 'component': '', 'monitor': 'localhost', 'agent': 'zencommand', 'severity': 0}

[29-Sep-2011 18:27:12] <rmatte> there's no mechanism where it checks dns for the new IP automatically

[29-Sep-2011 18:27:18] <jb> would that indicate an event should be populated?

[29-Sep-2011 18:27:52] <rmatte> jb: not necessarily

[29-Sep-2011 18:28:01] <rmatte> look at the severity

[29-Sep-2011 18:28:03] <rmatte> 0

[29-Sep-2011 18:28:06] <jb> ok, zencommand is reporting a success, but nothing is being populated.

[29-Sep-2011 18:28:06] <rmatte> that's a clear event

[29-Sep-2011 18:28:30] <jb> 2011-09-29 17:24:31,852 DEBUG zen.SshClient: 172.26.101.95 channel 1 CommandChannel closing command channel for command /usr/bin/sudo /usr/local/bin/check_bind.sh with data: "OK | 'success'=91712 'referral'=0 'nxrrset'=14907 'nxdomain'=24956 'recursion'=0 'failure'=0 'duplicate'=0 'dropped'=0\n"

[29-Sep-2011 18:28:30] <jb> 2011-09-29 17:24:31,852 DEBUG zen.SshClient: 172.26.101.95 channel 1 CommandChannel closing command channel for command /usr/bin/sudo /usr/local/bin/check_bind.sh with data: "OK | 'success'=91712 'referral'=0 'nxrrset'=14907 'nxdomain'=24956 'recursion'=0 'failure'=0 'duplicate'=0 'dropped'=0\n"

[29-Sep-2011 18:28:32] <jb> oops

[29-Sep-2011 18:28:33] <jb> sorry.

[29-Sep-2011 18:28:51] <rmatte> right, so that's working

[29-Sep-2011 18:29:04] <jb> i have a datapoint for each success, referral, etc.

[29-Sep-2011 18:29:14] <jb> it's set to auto though.. maybe I should change it to nagios?

[29-Sep-2011 18:29:17] <rmatte> not sure why your datapoints have single quotes around them

[29-Sep-2011 18:29:20] <tvincent> ok

[29-Sep-2011 18:29:23] <rmatte> leave it at auto

[29-Sep-2011 18:29:30] <jb> yeah i wondered if that is messing something up

[29-Sep-2011 18:29:32] <jb> the single quotes

[29-Sep-2011 18:30:13] <jb> let me try without them.

[29-Sep-2011 18:30:19] <rmatte> the format I'm used to seeing is: OK|success=91712 referral=0 nxrrset=14907 nxdomain=24956 recursion=0 failure=0 duplicate=0 dropped=0

[29-Sep-2011 18:30:59] <jb> ok

[29-Sep-2011 18:32:41] <jb> ok, removed those

[29-Sep-2011 18:32:45] <jb> there is also that \n at the end.,

[29-Sep-2011 18:32:52] <rmatte> the \n makes no difference

[29-Sep-2011 18:32:55] <jb> k

[29-Sep-2011 18:32:56] <rmatte> that's normal

[29-Sep-2011 18:33:07] <jb> 2011-09-29 17:32:18,051 DEBUG zen.RRDUtil: Creating new RRD file /opt/zenoss/perf/Devices/dns1.follett.com/bindStats_success.rrd

[29-Sep-2011 18:33:09] <jb> thats good

[29-Sep-2011 18:33:51] <rmatte> right, but did you see any for the others?

[29-Sep-2011 18:34:02] <rmatte> also, if you're using Zenoss 3.2, you did apply those 3 patches right?

[29-Sep-2011 18:34:45] <jb> this is 4.0.2

[29-Sep-2011 18:35:09] <rmatte> ah k, can't comment on 4.0.2 lol

[29-Sep-2011 18:35:28] <rmatte> anyways, you should see it create an RRD for each of those datapoints

[29-Sep-2011 18:36:07] <jb> hrm nope, nothing yhet

[29-Sep-2011 18:36:11] <jb> just one RRD

[29-Sep-2011 18:36:32] <jb> 011-09-29 17:35:14,512 DEBUG zen.SshClient: 172.26.101.95 channel 1 CommandChannel closing command channel for command /usr/bin/sudo /usr/local/bin/check_bind.sh with data: 'OK|success=94064 referral=0 nxrrset=15262 nxdomain=26113 recursion=0 failure=0 duplicate=0 dropped=0\n'

[29-Sep-2011 18:36:48] <jb> looks good to me.

[29-Sep-2011 18:36:51] <rmatte> then something is wrong with your actual Zenoss install because that makes no sense

[29-Sep-2011 18:37:02] <rmatte> you do actually have a datapoint created that matches each of those?

[29-Sep-2011 18:37:13] <jb> yessir

[29-Sep-2011 18:38:04] <rmatte> maybe command datasources are broken in 4.0.2 like they were in 3.2

[29-Sep-2011 18:38:12] * rmatte shrugs

[29-Sep-2011 18:39:05] <jb> hrm maybe so

[29-Sep-2011 18:39:39] <jb> http://thunder.jbdesign.net/~jbaird/zenoss/bindstats.png

[29-Sep-2011 18:40:23] <rmatte> ummm, I've never seen it shown like that, with the name of the datasource in front of the datapoint name

[29-Sep-2011 18:40:29] <rmatte> does it do that automatically or did you add that?

[29-Sep-2011 18:40:54] <jb> automatically

[29-Sep-2011 18:41:00] <rmatte> ah, weird, ok

[29-Sep-2011 18:41:14] <rmatte> but yeh, if you're only seeing one RRD created there's definitely an issue

[29-Sep-2011 18:41:24] <rmatte> the only thing I can suggest is change it from auto to nagios and see if that works

[29-Sep-2011 18:41:32] <rmatte> but auto should figure it out on it's own

[29-Sep-2011 18:41:37] <rmatte> I've never personally changed it

[29-Sep-2011 18:41:43] <jb> i'll try that

[29-Sep-2011 18:42:07] <Guest97443> rmatte: speaking of 3.2...  i m still using 3.0.3 wht u think abt 3.2??? can i use zenpacks those of 3.0.3 or so in 3.2??

[29-Sep-2011 18:42:30] <rmatte> Guest97443: yes, all it is is a ton of bugfixes on top of 3.0.3

[29-Sep-2011 18:42:41] <rmatte> there are no real new features

[29-Sep-2011 18:42:55] <jb> that fixed it!

[29-Sep-2011 18:42:56] <rmatte> if you use it, make sure you apply the patches listed in the channel topic

[29-Sep-2011 18:43:04] <jb> all RRD's created at least

[29-Sep-2011 18:43:04] <rmatte> 3.2 patches `zenpatch 26334`, `zenpatch 26335`, `zenpatch 26337`

[29-Sep-2011 18:43:18] <rmatte> jb: oh, if all the RRDs are created then you're good

[29-Sep-2011 18:43:26] <rmatte> should just be another 2 cycles before you see data

[29-Sep-2011 18:44:04] <Guest97443> ohh ok then i will run test on it... i think i can upgrade to 3.1 and then upgrade to 3.2??? and if i m correct i shouldnt be loosing my rrd files??

[29-Sep-2011 18:44:17] <Guest97443> and devices added should be existing?

[29-Sep-2011 18:44:32] <rmatte> Guest97443: yeh, it won't touch your RRDs

[29-Sep-2011 18:44:36] <rmatte> correct

[29-Sep-2011 18:44:48] <Guest97443> sweet

[29-Sep-2011 18:44:54] <rmatte> you don't lose data during upgrades unless something bad happens

[29-Sep-2011 18:45:11] <rmatte> you might want to backup your /usr/local/zenoss or /opt/zenoss directory just to be sure that

[29-Sep-2011 18:45:23] <rmatte> that way if something does go wrong you can just do a fresh install of that version then copy it back

[29-Sep-2011 18:45:27] <rmatte> but I doubt you'll have problems

[29-Sep-2011 18:45:33] <Guest97443> ohhh i hope it doesnt happens in my case :S

[29-Sep-2011 18:45:35] <Guest97443> ohh sure

[29-Sep-2011 18:45:48] <Guest97443> oh sure

[29-Sep-2011 18:45:50] <rmatte> I don't even know if you need to upgrade to 3.1 before going to 3.2

[29-Sep-2011 18:46:03] <rmatte> you might be able to just go from 3.0.3 straight to 3.2

[29-Sep-2011 18:46:09] <rmatte> check the upgrade paths in the install doc

[29-Sep-2011 18:47:49] <Guest97443> ohhh k... one more question is zenpacks, templates and mibs loaded will be existing or i hav to redo all??

[29-Sep-2011 18:47:56] <Guest97443> logically it should be

[29-Sep-2011 18:48:07] <Guest97443> as its an upgrade not a fresh install

[29-Sep-2011 18:48:17] <rmatte> you'll have everything you had before

[29-Sep-2011 18:48:22] <Guest97443> k cool

[29-Sep-2011 18:48:51] <Guest97443> thnks

[29-Sep-2011 18:56:44] <Guest97443> rmatte: how to display snmpwalk result as alert... for example. IPSEC current tunnels. I want to generate an alert from zenoss such a way tht if there is any change in IPSEC tunnel counts it should tigger an alert... and to show which tunnel came up or went down...for example if the current IPSEC tunnel count is say 30 and additional tunnel comes up then it should tigger an alert like 31 and info regarding 31st tunnel... vice versa

[29-Sep-2011 18:57:11] <Guest97443> [zenoss@zenoss root]$ snmpwalk -cpublic -On -v2c "devicename" 1.3.6.1.4.1.9.9.171.1.3.1.1.0 .1.3.6.1.4.1.9.9.171.1.3.1.1.0 = Gauge32: 9

[29-Sep-2011 19:05:30] <jb> i have to say.. i don't see that 4.0.2 is any faster

[29-Sep-2011 19:05:32] <Guest97443> i can do it as threshold but u know i hav to fix a limit for tht

[29-Sep-2011 19:05:32] <jb> the UI at least.

[29-Sep-2011 19:05:38] <jb> i think its slower than 3.1

[29-Sep-2011 20:07:57] kocolosk|away is now known as kocolosk

[30-Sep-2011 00:00:01] [disconnected at Fri Sep 30 00:00:01 2011]

[30-Sep-2011 00:00:02] [connected at Fri Sep 30 00:00:02 2011]

[30-Sep-2011 00:00:12] [zenoss-logger (logger bot) has joined #zenoss]

[30-Sep-2011 04:30:12] SDuensin_ is now known as SDuensin

[30-Sep-2011 04:30:12] kocolosk_ is now known as kocolosk

[30-Sep-2011 04:52:39] <ion-fan> what happened??? so much netsplits??

[30-Sep-2011 06:17:36] kocolosk is now known as kocolosk|away

[30-Sep-2011 08:30:58] <jmp242> morning all

[30-Sep-2011 10:01:47] <Hackman238> Morning all!

[30-Sep-2011 10:01:52] <jmp242> hi Hackman238

[30-Sep-2011 10:03:01] <Hackman238> jmp242: How goes it?

[30-Sep-2011 10:03:44] <jmp242> not too bad now that I've rebooted

[30-Sep-2011 10:04:28] <Hackman238> jmp242: LOL

[30-Sep-2011 10:07:58] <Hackman238> jmp242: Any idea what this guy is refering to? message/61648

[30-Sep-2011 10:21:41] hmp_ is now known as hmp

[30-Sep-2011 10:26:06] jb____ is now known as jb

[30-Sep-2011 10:26:14] <jb> morning all.

[30-Sep-2011 10:27:02] <dhopp> when looking at the event console there is a field for 'suppid'..what exactly is that used for?  It seems to be null for every event I have

[30-Sep-2011 10:27:53] <jb> bigegor: around?

[30-Sep-2011 10:30:15] <jb> hrm, I have a modeler that isn't properly populating the hardware model / os model / etc fields.

[30-Sep-2011 10:30:22] <jb> although the modeler is collecting this data properly.

[30-Sep-2011 10:34:05] <jmp242> Hackman238: yea

[30-Sep-2011 10:34:23] <jmp242> he's saying he doesn't understand Apache rewrite rules, but thinks Zenoss forums should help him with it

[30-Sep-2011 10:38:21] <rocket> jmp242: isnt that the rule of any support forum?

[30-Sep-2011 10:38:24] <rocket>

[30-Sep-2011 10:40:10] <dhopp> jmp242: what forum post are you referring to?

[30-Sep-2011 10:46:43] <jmp242>  message/61648

[30-Sep-2011 10:48:42] <SDuensin> FrIdaY!

[30-Sep-2011 10:49:02] <SDuensin> The CPU model doesn't have "cores"?

[30-Sep-2011 10:51:25] <jb> rocket: got the bind stuff working, mostly

[30-Sep-2011 10:51:29] <jb> rocket: thanks for the help.

[30-Sep-2011 10:56:59] <rocket> no problem .. its what I do .. all day .. every day .. for so long ...

[30-Sep-2011 10:57:57] <SDuensin> rocket, lucky you.  This thing hurts my head!

[30-Sep-2011 10:59:28] <rocket> SDuensin: I get the fun ones

[30-Sep-2011 11:00:13] <SDuensin> Then maybe you want to take over my Windows monitoring nightmare.  :-P

[30-Sep-2011 11:01:01] <jb> hrm, i dunno why this modeler isn't populating hardware model/os.

[30-Sep-2011 11:01:10] <rocket> SDuensin: are you an enterprise customer?

[30-Sep-2011 11:01:21] <SDuensin> rocket, sadly, no.

[30-Sep-2011 11:01:58] <jb> SDuensin: i have nightmares with windows monitoring as well.

[30-Sep-2011 11:02:02] <jb> ever since 3.x :/

[30-Sep-2011 11:02:42] <SDuensin> jb, I never could get WMI to work on anything.  So I'm doing it all myself over ssh.  It's working, but it's slow going since I have to look up how to do everything from other Zenoss ZenPacks.

[30-Sep-2011 11:02:53] <jb> windows via ssh?

[30-Sep-2011 11:02:55] <jb> eek.

[30-Sep-2011 11:03:00] <jb> WMI is full of problems.

[30-Sep-2011 11:03:07] <jb> memory leaks, stupid bugs, etc.

[30-Sep-2011 11:03:07] <SDuensin> Sounds bad, but it works.  And it works every time.

[30-Sep-2011 11:03:18] <jb> yeah but can you get perfmon counters with ssh?

[30-Sep-2011 11:03:43] <SDuensin> Now once I get there, I occasionally have to use WMI or SNMP - however, it's as an admin on the box I need the data from.  No security hassles or other stupidity.

[30-Sep-2011 11:19:50] <ion-fan> Hackman238: did you get further with the zenpack?

[30-Sep-2011 11:53:14] <davetoo> wow

[30-Sep-2011 11:53:16] <davetoo> crickets

[30-Sep-2011 11:55:27] <jmp242> lunchtime

[30-Sep-2011 12:00:01] [disconnected at Fri Sep 30 12:00:01 2011]

[30-Sep-2011 12:00:01] [connected at Fri Sep 30 12:00:01 2011]

[30-Sep-2011 12:00:13] [zenoss-logger (logger bot) has joined #zenoss]

[30-Sep-2011 12:08:00] <nyeates> lunchtime indeed

[30-Sep-2011 12:08:30] <nyeates> anyone had any experiences with 3.2.1 beta?

[30-Sep-2011 12:08:43] <nyeates> the good, the bad, the ugly... the religious?

[30-Sep-2011 12:08:50] <davetoo> when I I'm actually fully awake for my daily 07:30 meeting, by now (09:00) it seems nearly lunchtime.

[30-Sep-2011 12:10:56] <nyeates> wow its only 9 there

[30-Sep-2011 12:10:57] <nyeates> crazy

[30-Sep-2011 12:12:08] <nyeates> btw, #bash and #git have been life savers

[30-Sep-2011 12:12:19] <nyeates> i had no idea how ative and helpful irc chats like that would be

[30-Sep-2011 12:12:46] <davetoo> http://tldp.org/LDP/abs/html/  <-- you have seen that by now, I hope?

[30-Sep-2011 12:13:31] <davetoo> and yes, Freenode is not EFNet

[30-Sep-2011 12:14:26] <nyeates> i do not have that link

[30-Sep-2011 12:14:31] <nyeates> fml

[30-Sep-2011 12:14:41] <nyeates> looks very helpful....there is a lot out there tho

[30-Sep-2011 13:09:59] kocolosk|away is now known as kocolosk

[30-Sep-2011 13:14:08] bigegor_ is now known as bigegor

[30-Sep-2011 13:22:49] <jb> hmm

[30-Sep-2011 13:22:57] <jb> ok, am I being stupid or is this graph totally wrong.

[30-Sep-2011 13:23:06] <jb> http://thunder.jbdesign.net/~jbaird/zenoss/queries.png

[30-Sep-2011 13:23:29] <jb> its showing 100+ success queries on the graph, but the legend says the max is 40.

[30-Sep-2011 13:29:32] <jmp242> the graph is stacked

[30-Sep-2011 13:29:38] <jmp242> add up each of the colors from the legend

[30-Sep-2011 13:29:44] <jmp242> you'll see they add up to ~ 102

[30-Sep-2011 13:30:41] <jb> oh thats the problem.

[30-Sep-2011 13:30:46] <jb> hah, the datapoints are stacked.

[30-Sep-2011 13:30:49] <jb> argh.  thanks

[30-Sep-2011 13:31:28] <jb> i guess I was unclear as to what "stacked" meant.

[30-Sep-2011 13:31:30] <jmp242> np

[30-Sep-2011 13:31:45] <jb> when would you want to use stacked?

[30-Sep-2011 13:32:01] <Omeganon> this is a great example of why

[30-Sep-2011 13:32:11] <jmp242> well, if you didn't you wouldn't see the smaller values?

[30-Sep-2011 13:32:22] <Omeganon> you have a lot of individual subsets of a particular category where you care about the subset counts as well as the total of all of them.

[30-Sep-2011 13:32:57] <jb> hrm, yeah i can't even make sense of this graph if the points are not stacked.

[30-Sep-2011 13:33:17] <jb> all of the colors blend together

[30-Sep-2011 13:33:27] <Omeganon> it's also easier to see relative magnitude of the subsets

[30-Sep-2011 13:34:02] <jb> hm

[30-Sep-2011 13:34:11] <jb> so really, you would want to stack the "smaller" subsets/values

[30-Sep-2011 13:34:55] <Omeganon> I think the png you showed is the right way to do it.

[30-Sep-2011 13:35:11] <jb> oh

[30-Sep-2011 13:35:19] <Omeganon> stack them all

[30-Sep-2011 13:36:01] <jb> ok, so basically you just need to interpret the graph differently

[30-Sep-2011 13:38:15] <Omeganon> if you're not used to that kind of view. The most important thing to take from that graph is how many total queries/s is your server doing. That's going to be the best metric as to how busy the server is. From there, you can easily see from the stack what are the biggest consumers of those queries. In yours, it appears to be lookups for non-existent domains followed closely by successful lookups.

[30-Sep-2011 13:38:43] <Omeganon> From there, I'd look at my server logs to see what host(s) are sending all those requests for non-existent domains; I'd suspect they had malware on them

[30-Sep-2011 13:40:03] <Omeganon> The stacked view gives you the best of both types of views.

[30-Sep-2011 13:43:47] <jb> Omeganon: very good point.

[30-Sep-2011 13:43:48] <jb> thanks!

[30-Sep-2011 13:44:17] <jb> just wrapping up the BIND9 pack

[30-Sep-2011 13:44:32] <jb> it has that query overview graph, a query RTT graph, zones, recursive clients

[30-Sep-2011 13:44:42] <jb> might add detailed query types

[30-Sep-2011 13:44:48] <jb> but parsing it is a pain :/

[30-Sep-2011 13:53:04] <bigegor> hi all

[30-Sep-2011 13:55:43] <bigegor> jb: you want to ask something?

[30-Sep-2011 14:22:16] <jb> hey egor

[30-Sep-2011 14:22:18] <jb> yeah

[30-Sep-2011 14:22:19] <jb> still here?

[30-Sep-2011 14:24:57] kocolosk is now known as kocolosk|away

[30-Sep-2011 14:27:18] otakup0p1 is now known as otakup0pe

[30-Sep-2011 14:37:35] kocolosk|away is now known as kocolosk

[30-Sep-2011 14:38:03] <bigegor> yes

[30-Sep-2011 14:50:50] <jb> k

[30-Sep-2011 14:50:53] <jb> i sent you some msgs.

[30-Sep-2011 14:51:14] <jb> bigegor: for some reason, on my 4.0.2 install, DellMon is no longer changing the Hardware Model / OS version on the "Overview" page.

[30-Sep-2011 14:51:22] <jb> however, your modeler is picking this info up.

[30-Sep-2011 14:51:29] <jb> 2011-09-30 09:25:26,276 DEBUG zen.ZenModeler: Plugin community.snmp.DellDeviceMap results =

[30-Sep-2011 14:51:29] <jb>       ({'.1.3.6.1.4.1.674.10892.1.300.10.1.9.1': 'PowerEdge R610', '.1.3.6.1.4.1.674.10892.1.400.10.1.6.1': 'Red

[30-Sep-2011 14:51:30] <jb>       Hat Enterprise Linux Server', '.1.3.6.1.4.1.674.10892.1.300.10.1.11.1': 'BN4DSL1'}, {})

[30-Sep-2011 14:51:47] <jb> oops, sorry for the multiline.

[30-Sep-2011 14:52:18] <jb> so, Hardware Model is showing as "Net-SNMP" and the OS is showing as generic.

[30-Sep-2011 14:54:05] <bigegor> sorry, no idea.

[30-Sep-2011 14:54:28] <jb> hm

[30-Sep-2011 14:58:29] <jb> well, adding zen's own DellDeviceMap back into the modeler list fixed it.

[30-Sep-2011 14:58:39] <jb> for some reason, your DellDeviceMap isn't doing the trick.

[30-Sep-2011 15:05:02] <bigegor> sorry, i can't test my ZenPacks against Zenoss Ent 4.0.2

[30-Sep-2011 15:05:59] <jmp242> jb: wait for Core 4.2

[30-Sep-2011 15:06:04] <jmp242> then Egor can test the zenpacks

[30-Sep-2011 15:09:04] <zirikili> hi guys

[30-Sep-2011 15:09:33] <zirikili> Hackman238:  hello

[30-Sep-2011 15:10:20] <bigegor> bye

[30-Sep-2011 15:13:39] <jb> yep, i know.  thanks

[30-Sep-2011 15:14:32] <jb> hm.  what do I need to do to be able to check the "Monitored" box under a devices IP Services

[30-Sep-2011 15:14:39] <jb> right now, the box is uncheckable.

[30-Sep-2011 15:17:55] <jmp242> There was a patch for 3.1 for that

[30-Sep-2011 15:18:02] <jmp242> IDK if it would apply to 4.x

[30-Sep-2011 15:18:20] <jmp242> anyway, time for leaving - have a good weekend everyone!

[30-Sep-2011 15:31:44] <SDuensin> Are the "Memory/Swap" fields on the device overview modeler fields or performance fields?

[30-Sep-2011 15:31:52] * SDuensin is slowly working this all out.

[30-Sep-2011 15:43:16] <zirikili> I don't know what is wrong. I have a device which is down (not responding) but in zenoss there is no flag warning it is down. it is a /Ping device, by the way. Any help ?

[30-Sep-2011 15:43:42] <rocket> zirikili: what are the zenping logs showing you for the device?

[30-Sep-2011 15:44:07] <zirikili> rocket: could you tell me where can I check it ?

[30-Sep-2011 15:44:23] <rocket> cd $ZENHOMe

[30-Sep-2011 15:44:28] <rocket> find . -name zenping.log

[30-Sep-2011 15:47:05] <zirikili> zenping.log is empty  !!!

[30-Sep-2011 15:47:58] <rocket> is zenping running?

[30-Sep-2011 15:48:42] <zirikili> oops

[30-Sep-2011 15:48:51] <zirikili> is not. how do I start it?

[30-Sep-2011 15:49:08] <rocket> zenping run

[30-Sep-2011 15:49:19] <rocket> err zenping start

[30-Sep-2011 15:50:26] <zirikili> should I use the zenoss user?

[30-Sep-2011 15:50:39] <rocket> yes

[30-Sep-2011 15:50:46] <dhopp> zirikili:  yes…almost everything in zenoss should be done as the zenoss user

[30-Sep-2011 15:50:47] <rocket> or as the zenoss user

[30-Sep-2011 15:50:51] <rocket> zenoss status

[30-Sep-2011 15:50:54] <rocket> zenoss start

[30-Sep-2011 15:54:26] <zirikili> well… after  zenoss start   the zenping and zensyslog was staterd because they was not running.  but when i run zenoss status that two process are still "not running"  . the other daemons seams ok.

[30-Sep-2011 15:55:59] <zirikili> *seems

[30-Sep-2011 15:56:30] <dhopp> zirikili:  can you do zenping run -v10 (as zenoss user)

[30-Sep-2011 15:56:47] <dhopp> zirikili:  it will run through and then stop immediately..but I'm curious if it throws a traceback

[30-Sep-2011 15:59:08] <zirikili> it shows some  python debug errors and the last line is "OSError: [Errno 13] Permission denied"

[30-Sep-2011 15:59:21] <dhopp> zirikili:  what's the line above permission denied?

[30-Sep-2011 15:59:56] <rocket> he has zensocket setuid permissions  issues

[30-Sep-2011 16:00:08] <zirikili> " File "/usr/local/zenoss/python/lib/python2.6/os.py", line 368, in _execvpe

[30-Sep-2011 16:00:08] <zirikili>     func(file, *argrest) "

[30-Sep-2011 16:00:39] <rocket> post the whole traceback in a pastebin please

[30-Sep-2011 16:01:05] <dhopp> zirikili:  yes please post the traceback to pastebin...

[30-Sep-2011 16:01:22] <dhopp> zirikili:  also, can you do ls -l $ZENHOME/bin/zensocket

[30-Sep-2011 16:04:47] <zirikili> ok… it is here:   http://pastebin.com/N4u4wQdr

[30-Sep-2011 16:08:40] <dhopp> zirikili:  what does ls -ld $ZENHOME/log/ and ls -l $ZENHOME/log/zenping.log look like?

[30-Sep-2011 16:09:22] <zirikili> $ ls -ld $ZENHOME/log/

[30-Sep-2011 16:09:22] <zirikili> drwxr-xr-x 2 zenoss zenoss 4096 Sep 28 09:05 /usr/local/zenoss/zenoss/log/

[30-Sep-2011 16:09:55] <zirikili> $ ls -l $ZENHOME/log/zenping.log

[30-Sep-2011 16:09:55] <zirikili> -rw-r--r-- 1 zenoss zenoss 0 Apr 10 23:07 /usr/local/zenoss/zenoss/log/zenping.log

[30-Sep-2011 16:10:30] <dhopp> zirikili:  is this a new install?

[30-Sep-2011 16:10:43] <zirikili> yes

[30-Sep-2011 16:10:52] <dhopp> zirikili:  how did you install it?

[30-Sep-2011 16:11:10] <zirikili> apt-get  on a debian box

[30-Sep-2011 16:12:50] <dhopp> zirikili:  I'm going to have to defer to rocket…the things that I have seen cause this type of error seem ok

[30-Sep-2011 16:14:25] <zirikili> do you recommend any specific linux distro for zenoss or anyone could work fine?

[30-Sep-2011 16:15:29] <dhopp> zirikili:  I personally use redhat or centos..but there are people that use ubuntu and suse

[30-Sep-2011 16:16:09] <dhopp> zirikili:  can you look in /etc/sudoers …is there any entries for zenoss?

[30-Sep-2011 16:17:20] <dhopp> although I don't think that's necessarily the problem because my dev box doesn't have any sudo entries and it works fine

[30-Sep-2011 16:18:07] <rocket> this box had to have been installed incorrectly or something because its trying to open the priviledged port for pinging and it cant do that

[30-Sep-2011 16:18:29] <rocket> zenping and zensyslog both use zensocket to do this

[30-Sep-2011 16:18:32] <jb> hrm i guess 4.0.2 has the same bug that 3.2 did in regards to IP service monitoring.

[30-Sep-2011 16:19:20] <rocket> what are the contents of zensocket?

[30-Sep-2011 16:19:22] <dhopp> jb: which is?

[30-Sep-2011 16:19:27] <zirikili> ok… there is no zenoss entry on sudoers

[30-Sep-2011 16:19:46] <rocket> your filesize for zensocket is wrong

[30-Sep-2011 16:20:09] <dhopp> rocket:  doh!  I missed that

[30-Sep-2011 16:20:10] <rocket> dont look at the contents I guess .. but basically I think your zensocket file is corrupt

[30-Sep-2011 16:20:31] <dhopp> zirikili: -rwsr-x--- 1 root zenoss 9720 Aug 25 22:08 /opt/zenoss/bin/zensocket

[30-Sep-2011 16:20:36] <dhopp> yours is 0 byte

[30-Sep-2011 16:20:54] <dhopp> (I just pasted what mine looks like)

[30-Sep-2011 16:21:07] <rocket> my zensocket file is only 9720  yours is 37827?

[30-Sep-2011 16:21:09] <zirikili> hey… I just changed the suid and root user for zensocket...

[30-Sep-2011 16:21:20] <zirikili> it is working now

[30-Sep-2011 16:21:39] <jb> dhopp: that you can't enable monitoring of the IP services.

[30-Sep-2011 16:21:43] <dhopp> zirikili: your isn't 0 byte..I was looking at your zenping.log file..oops!  it's been a long day

[30-Sep-2011 16:22:40] <dhopp> jb: I think there was a patch for 3.x to fix that…not sure if there is a patch for 4.x enterprise...

[30-Sep-2011 16:26:44] <dhopp> jb: I can enable/disable IP services on my 3.2 install

[30-Sep-2011 16:34:03] <zirikili> the zenping is working fine now …. thank you  rocket and dhopp… very much

[30-Sep-2011 16:35:42] <SDuensin> Ok, things like this hurt my head...  Why is there a getDeviceName but no setDeviceName?

[30-Sep-2011 16:36:10] <dhopp> SDuensin:  because that would be too obvious :-P

[30-Sep-2011 16:36:43] <SDuensin> I'm writing some modeler plugins and want to set the "Device Name" to the SNMP SysName.

[30-Sep-2011 16:37:00] <SDuensin> I also want to set the Uptime.  Can't figure that out either.  :-(

[30-Sep-2011 16:38:56] <dhopp> SDuensin: I *think* you just want to set dev.id

[30-Sep-2011 16:39:45] <SDuensin> Can I do that from inside the "process" method?  Just set it?  So far, everything I've set in a modeler has been returned in some kind of map.

[30-Sep-2011 16:40:16] <dhopp> SDuensin:  www.shanewilliamscott.com has a modeler that does what you want…you could probably look at what he did and incorporate it into what you want

[30-Sep-2011 16:40:28] <SDuensin> Yea.  Did that.  :-P

[30-Sep-2011 16:40:38] <SDuensin> Lemme look at it again.  I know more now than I did before.   :-D

[30-Sep-2011 16:40:46] <dhopp> SDuensin:  ok…I haven't dug into writing custom modelers yet so you are ahead of me

[30-Sep-2011 16:42:36] <SDuensin> Python is not my strong point (yet).  This is all uphill.

[30-Sep-2011 16:44:44] <SDuensin> Yea.  Didn't work.

[30-Sep-2011 16:45:47] <SDuensin> All this DeviceMap and SNMPMap and stuff still confuses me.

[30-Sep-2011 16:46:53] <dhopp> hmm apparently zenhub did NOT like me stopping mysql for a short time

[30-Sep-2011 16:46:53] <dhopp> heh

[30-Sep-2011 16:57:21] kocolosk is now known as kocolosk|away

[30-Sep-2011 16:59:55] kocolosk|away is now known as kocolosk

[30-Sep-2011 17:25:35] >dhopp> hmm..it seems that the "Printable" link on a graph report always shows the hourly graphs...

[30-Sep-2011 17:26:09] kocolosk is now known as kocolosk|away

[30-Sep-2011 18:01:16] Diddi_ is now known as Diddi

[30-Sep-2011 18:33:39] kocolosk|away is now known as kocolosk

[30-Sep-2011 20:47:34] kocolosk is now known as kocolosk|away

[30-Sep-2011 21:25:07] kocolosk|away is now known as kocolosk

[30-Sep-2011 22:06:16] kocolosk is now known as kocolosk|away

[30-Sep-2011 23:39:30] kocolosk|away is now known as kocolosk

[30-Sep-2011 23:53:15] kocolosk is now known as kocolosk|away

[30-Sep-2011 23:53:26] kocolosk|away is now known as kocolosk

[30-Sep-2011 23:57:30] kocolosk is now known as kocolosk|away

[01-Oct-2011 00:00:01] [disconnected at Sat Oct  1 00:00:01 2011]

Comments (0)