[03-Feb-2011 10:54:18] <Sam-I-Am> brrrrr its cold out
[03-Feb-2011 10:54:59] <Sam-I-Am> nyeates: howdy
[03-Feb-2011 10:55:45] <rm-rf> nyeates: thanks. i just added 'zenoss.snmp.UCDHardDiskMap' to the modeler plugins that this particular system is using
[03-Feb-2011 10:55:50] <nyeates> hola sam
[03-Feb-2011 10:56:16] <rm-rf> that is getting me closer, but it appears to only be picking up /dev/sda, and nothing else
[03-Feb-2011 10:57:00] <nyeates> rm-rf: look at some of the Configuration Properties (zProperties). There are a few that block certain mounts (such as usb or cd mounts)
[03-Feb-2011 10:57:04] <nyeates> maybe its blocking other mounts too
[03-Feb-2011 10:57:36] pmcguire is now known as ptmcg
[03-Feb-2011 10:59:55] <nyeates> Paul, do you know anything about the python side of zenproperties?
[03-Feb-2011 11:00:15] <nyeates> ...someone was asking some questions i couldnt answer.
[03-Feb-2011 11:00:41] <ptmcg> I could make up something that sounds convincing, but you probably want actual information
[03-Feb-2011 11:01:40] <ptmcg> I know that the zenprops are evaluated down the class inheritance chain, so that a property defined at /Server/Windows will override one defined at /Server
[03-Feb-2011 11:02:04] <nyeates> [10:39]  <davetoo> ZenPropertyManager does not override everything from PropertyManager so somethings seem broken
[03-Feb-2011 11:02:12] <ptmcg> and I believe Zope acquisition is the mechanism used for this, but now I've moved into the speculation area
[03-Feb-2011 11:02:34] <kells> Lemme check that for you
[03-Feb-2011 11:02:34] <nyeates> <davetoo> for example, dmd.zenPropertyItems() throws: AttributeError: dmdRootName
[03-Feb-2011 11:03:12] <nyeates> davetoo> but, of course, dmd.Devices.zenPropertyItems() works fine.
[03-Feb-2011 11:03:14] <kells> Why would you do that from there?  Almost everything IIRC is done from dmd.Devices
[03-Feb-2011 11:03:21] <kells> Yeah.
[03-Feb-2011 11:03:35] <kells> That's roughly what I would expect
[03-Feb-2011 11:03:39] <davetoo> I'm writing a unit test suite to figure it out 
[03-Feb-2011 11:03:52] <ptmcg> ah, this actually sounds similar to a bug that was I just fixed last week
[03-Feb-2011 11:03:55] * ptmcg checking...
[03-Feb-2011 11:03:59] <kells> There are different properties for different areas, so dmd.Networks has its own
[03-Feb-2011 11:04:45] <davetoo> kells: it causes MigrateUtils.migratePropertyType() to fail
[03-Feb-2011 11:05:04] <ptmcg> Check out this ticket - is this related to your question?  http://dev.zenoss.com/trac/ticket/5464
[03-Feb-2011 11:05:15] <davetoo> if you try to run _getInstanceValues on dmd
[03-Feb-2011 11:05:31] <davetoo> ptmcg: no.  That's my ticket, anyway   And it's different.
[03-Feb-2011 11:05:41] <davetoo> davetoo = dcarmean
[03-Feb-2011 11:05:56] <davetoo> kells: Principle of Least Surprise.
[03-Feb-2011 11:06:16] <ptmcg> Your secret identity?! It never dawned on me that you two are never in the room at the same time...
[03-Feb-2011 11:07:25] <kells> Wasn't that changeset the one that broke the build for a day?
[03-Feb-2011 11:07:33] <ptmcg> FYI #5464 has been patched and verified thru QA, so it will be in the 3.1 release
[03-Feb-2011 11:07:37] <nyeates> trouble is, everyone considers different things as 'surprising'
[03-Feb-2011 11:07:44] <ptmcg> Kells, you are such a kidder!!!
[03-Feb-2011 11:07:56] <ptmcg> I don't think so
[03-Feb-2011 11:08:34] <kells> http://dev.zenoss.org/trac/changeset?new=25431%40sandboxen%2Fjhanson%2Fdevtypes-migrate-script&old=25413%40trunk
[03-Feb-2011 11:08:59] <davetoo> dmd is the top object that is an instance of (the mix-in) ZenPropertyManager.
[03-Feb-2011 11:09:12] <jrh0090> Yeah, having the default as [('','')] was causing all sorts of headaches
[03-Feb-2011 11:09:20] <davetoo> Thus, the Principle of Least Surprise makes me expect that any method in ZenPropertyManager should work.
[03-Feb-2011 11:09:23] <jrh0090> so the easiest thing was to make it an empty array
[03-Feb-2011 11:09:31] <davetoo> jrh0090: oh?  What headaches?
[03-Feb-2011 11:10:02] <ptmcg> When you say "array" you mean "list", of course
[03-Feb-2011 11:10:05] <davetoo> because if so, I'm breaking my system 
[03-Feb-2011 11:10:18] <jrh0090> if you look at DeviceClass.register_devtype it was appending to the new tuple to the blank one
[03-Feb-2011 11:10:34] <jrh0090> so you have [('',''), ('foo', 'bar')] when you registered  a devtype
[03-Feb-2011 11:10:39] <jrh0090> which is probably not what you would want
[03-Feb-2011 11:10:41] <rm-rf> any idea why zenoss would be reporting on /dev/sda3 after adding the UCDHardDiskMap plugin to the modeler, when the server doesn't actually have /dev/sda3?
[03-Feb-2011 11:10:42] <davetoo> oh
[03-Feb-2011 11:11:05] <rm-rf> oh, durr. /dev/sda3 == swap
[03-Feb-2011 11:11:05] <jrh0090> so making it an empty list allowed the property to always be there and actually be meaningful if something was presetn
[03-Feb-2011 11:12:27] <nyeates> rm-rf: check out 'zFileSystemMapIgnoreTypes'
[03-Feb-2011 11:12:41] <davetoo> I'm glad I was there, then, because I was patching zenoss_init_post to shove that into zendmd on first run
[03-Feb-2011 11:12:48] <nyeates> but yeah, yer swap and other various file systems may show up
[03-Feb-2011 11:13:14] <davetoo> but..
[03-Feb-2011 11:13:29] <davetoo> is it dmd.Devices.registerDevType(,) or ((,))?
[03-Feb-2011 11:13:56] <jrh0090> dmd.Devices.registerDevType('Name', 'Description')  I believe
[03-Feb-2011 11:14:15] <davetoo> ok, hang no,
[03-Feb-2011 11:14:17] <jrh0090> def register_devtype(self, description, protocol):
[03-Feb-2011 11:14:19] <jrh0090> i mean
[03-Feb-2011 11:14:46] <davetoo> Is it only correct in somebody's sandbox right now, then?
[03-Feb-2011 11:15:05] <jrh0090> nope, that has been patched to the 3.1 release
[03-Feb-2011 11:15:16] <jrh0090> Kells just linked to my sandbox to call me out :)_
[03-Feb-2011 11:15:28] <davetoo> dmd.Devices.register_devtype(,)  throws a SyntaxError
[03-Feb-2011 11:15:43] <davetoo> so does dmd.Devices.register_devtype((,))
[03-Feb-2011 11:15:43] <ptmcg> Well, he was calling *me* out, actually, since I had botched the first fix
[03-Feb-2011 11:15:53] <jrh0090> dmd.Devices.register_devtype('', '') will do what you want i believe
[03-Feb-2011 11:16:10] <davetoo> but I thought you said that this is what created the problem?
[03-Feb-2011 11:16:24] <kells> More like "Oh hey!  Now I think I know what you guys are talking about." 
[03-Feb-2011 11:16:42] <kells> Didn't mean to put anybody on the spot, I just have a poor memory.
[03-Feb-2011 11:16:43] <davetoo> jrh0090: ?
[03-Feb-2011 11:16:59] <davetoo> what exactly was the "fix" that broke things?
[03-Feb-2011 11:17:37] <jrh0090> adding the default to the top level device class didn't really break things, we just had to make sure that our implementation was aware of the difference
[03-Feb-2011 11:17:40] <jrh0090> if that makes sense 
[03-Feb-2011 11:17:42] <davetoo> that the first attempt was actually an (untested?) syntax error, the empty tuple?  And that
[03-Feb-2011 11:18:06] <davetoo> a 2-tuple of null strings is the correct answer?
[03-Feb-2011 11:18:30] <jrh0090> the only place we use the devtypes (that I could find) was the drop down list of "Protocols" for new devices
[03-Feb-2011 11:18:46] <davetoo> That's all I found, as well.
[03-Feb-2011 11:18:47] <kells> "Much Ado About Nothing"
[03-Feb-2011 11:18:57] <kells> Shakespeare is everywhere. 
[03-Feb-2011 11:19:17] <jrh0090> and the tuple of empty strings was just causing data to be missing from that list
[03-Feb-2011 11:19:36] <jrh0090> it really wasn't a big deal to change, fortunately we have a pretty good QA department that noticed something was missing from it
[03-Feb-2011 11:19:52] <davetoo> well, this came about because I have a tool that walks down the tree and writes properties to a json structure,
[03-Feb-2011 11:19:56] <ptmcg> davetoo: this is the original changeset: http://dev.zenoss.com/trac/changeset/25336
[03-Feb-2011 11:20:10] <rm-rf> how long does it typically take for zenoss to start graphing results?
[03-Feb-2011 11:20:41] <ptmcg> but jrh0090's fix cleans things up
[03-Feb-2011 11:20:57] <nyeates> rm-rf: a couple monitoring cycles... 3 technically, but sometimes u cant see it clearly until you get more
[03-Feb-2011 11:20:58] <davetoo> which is the "good" changeset?
[03-Feb-2011 11:21:32] <davetoo> I would prefer to use zenpatch than my own hack
[03-Feb-2011 11:21:39] <Jane_Curry> Hiya
[03-Feb-2011 11:21:47] <davetoo> oh, I can't,
[03-Feb-2011 11:22:24] <jrh0090> I think ptmcg is using his trac-fu to find it
[03-Feb-2011 11:22:28] <davetoo> since I'm still on 3.0.x, that step won't run.  and, n/mind, it's in 25340.  I'm still half asleep.
[03-Feb-2011 11:22:43] <nyeates> rm-rf: look in advanced > collectors > <collectorname> > overview > SNMP Performance Cycle Interval
[03-Feb-2011 11:22:48] <davetoo> oh, wait a second...
[03-Feb-2011 11:22:55] <nyeates> rm-rf default is 300 sec
[03-Feb-2011 11:23:15] <nyeates> Hi jane
[03-Feb-2011 11:23:18] <davetoo> 25340 is still wrong, yeah?
[03-Feb-2011 11:24:02] <davetoo> 25432
[03-Feb-2011 11:24:45] <davetoo> http://dev.zenoss.com/trac/changeset/25432  seems to be the fix
[03-Feb-2011 11:25:34] <rmatte> oooh, just noticed the beta is out
[03-Feb-2011 11:25:42] <rmatte> is the link on the download page?
[03-Feb-2011 11:25:49] <ptmcg> http://dev.zenoss.com/trac/changeset?new=25432%40trunk%2FProducts%2FZenModel&old=25410%40trunk%2FProducts%2FZenModel
[03-Feb-2011 11:26:02] <ptmcg> oh, you found it already
[03-Feb-2011 11:26:38] <davetoo> yes, thanks
[03-Feb-2011 11:27:10] <ptmcg> That fixes this change: http://dev.zenoss.com/trac/changeset?new=25336%40trunk%2FProducts%2FZenModel&old=25329%40trunk%2FProducts%2FZenModel
[03-Feb-2011 11:27:36] <davetoo> I'm glad I was in here; my hack to shove a HEREDOC into zendmd in zenoss_init_post was doing it the "wrong" way.
[03-Feb-2011 11:27:59] <rmatte> eugh, looks like I'll be re-logging the nmap plugin issue as a new ticket
[03-Feb-2011 11:28:02] <rmatte> since they decided to close it
[03-Feb-2011 11:28:12] <davetoo> rmatte: what's the bug?
[03-Feb-2011 11:28:37] <kells> davetoo: zendmd --script=filename   wasn't working for you?  Maybe I'm not understanding the HEREDOC comment properly
[03-Feb-2011 11:28:45] <rmatte> davetoo: the nmap plugin for detecting ip services to monitor hasn't worked for as long as I remember
[03-Feb-2011 11:29:03] <rmatte> and there were steps in the ticket to fix it, which I put there, but no one accurately attempted them
[03-Feb-2011 11:29:13] <rmatte> I'm going to put a formal patch together, test it, and re-submit
[03-Feb-2011 11:30:17] <davetoo> kells: ${ZENHOME}/bin/zendmd <<EOT
[03-Feb-2011 11:30:17] <davetoo> if not dmd.Devices.hasProperty('devtypes'):  dmd.Devices.register_devtype('','')
[03-Feb-2011 11:30:18] <davetoo> if not dmd.Devices.Server.hasProperty('devtypes'): dmd.Devices.Server.register_devtype('','')
[03-Feb-2011 11:30:18] <davetoo> dmd.Devices.cDateTest = ''
[03-Feb-2011 11:30:18] <davetoo> dmd.Systems.setZenProperty('systemClass', '')
[03-Feb-2011 11:30:18] <davetoo> commit()
[03-Feb-2011 11:30:18] <davetoo> EOT
[03-Feb-2011 11:30:20] <davetoo> (sorry)
[03-Feb-2011 11:30:57] <davetoo> I put that at the end of ${ZENHOME}/bin/zenoss_init_post
[03-Feb-2011 11:32:23] <davetoo> but now I should be able to use zenpatch
[03-Feb-2011 11:33:18] <davetoo> which I will test right now.  Hurray for VM snapshots 
[03-Feb-2011 11:33:51] <davetoo> now, about that beta 
[03-Feb-2011 11:39:42] <nyeates> 3.1.0 beta found at http://alpha.zenoss.com/files/
[03-Feb-2011 11:39:54] <nyeates> blogs/zenossblog/2011/02/01/zenoss-310-beta-is-now-available
[03-Feb-2011 11:39:56] <nyeates>  for details
[03-Feb-2011 11:40:05] <Jane_Curry> Anyone got any good answers to a blank component display in 3.0.3
[03-Feb-2011 11:41:07] <nyeates> which display jane?
[03-Feb-2011 11:41:38] <Jane_Curry> The Firefox Console gives "this.config[a] is undefined"
[03-Feb-2011 11:42:03] <Jane_Curry> This is a new ZenPack that started from my BridgeMIB ZenPack code
[03-Feb-2011 11:43:34] <Jane_Curry> So - new device type, with a new component.  Component shows in left menu but clicking component gives resulting blank ComponentGridPanel
[03-Feb-2011 11:46:18] <nyeates> Jane: i dont think i will be able to help here besidse to clarify - are there sub-components defined in this new device type?
[03-Feb-2011 11:47:43] <Sam-I-Am> Jane_Curry: btw your event management paper is awesome, thanks
[03-Feb-2011 11:48:35] <nyeates> like on a normal device, there might be Components > and below that "Interfaces", "Network Routes", "Stack Modules" .... does your device type have various types of Components listed? Or just one? or none?
[03-Feb-2011 11:50:20] <Jane_Curry> I guess I have a subcomponent in your terminology - so under Components on the LH menu, I see my new subcomponent type
[03-Feb-2011 11:51:09] <Jane_Curry> I van also use the ZMI and see the new device, its new relationship and the detils of its new component
[03-Feb-2011 11:51:56] <Jane_Curry> It's just the detail in the GUI that is blank - I should se 1 line for each of the instances of my new component
[03-Feb-2011 11:52:33] <Jane_Curry> and the Display drop down should give Details, Graphs etc but they are all blank
[03-Feb-2011 11:53:25] <Jane_Curry> <Sam-I-Am> - thanks.  We are wondering about offering the associated 3-day workshop as a remote class offering..
[03-Feb-2011 11:54:09] <Jane_Curry> .. probably too late for you if you have digested the paper - but I'd be interested to see whether there is interest in such a project??
[03-Feb-2011 11:57:31] <Sam-I-Am> Jane_Curry: i'd surely be interested... its just whether or not management here would go for it.
[03-Feb-2011 11:58:13] <Sam-I-Am> Jane_Curry: other topics would be interesting too
[03-Feb-2011 12:02:54] <nyeates> Jane: I cant give any feedback on the missing component instances and graphs besides that: you may want to browse newish trac tickets for issues devs had with this area...it changed a lot for 3.0 and there might be hints at issues
[03-Feb-2011 12:05:25] <Jane_Curry> <Sam-I-Am> We have just advertised the 1-day FastStart Workshop but this is definitely an intro workshop
[03-Feb-2011 12:05:51] <Jane_Curry> We thought we'd start with the short one.  I've run it several times as a face-to-face workshop
[03-Feb-2011 12:06:02] <Jane_Curry> but we haven't tried the remote format before
[03-Feb-2011 12:06:50] <nyeates> i think u might see more willingness to attend
[03-Feb-2011 12:07:01] <nyeates> cause now it will be open to those in other continents
[03-Feb-2011 12:07:46] <Sam-I-Am> remote would really help
[03-Feb-2011 12:07:48] <Jane_Curry> That's the thinking - that's partly why we deliberately haven't advertised a time on the workshop announcement
[03-Feb-2011 12:07:53] <Sam-I-Am> i attend lots of remote stuff
[03-Feb-2011 12:08:03] <nyeates> This concludes the developer session for this week. I encourage all of you to take 3.1.0 beta for a spin and to let us know how it goes. Our engineering is looking for feedback.
[03-Feb-2011 12:08:37] <Jane_Curry> But I've hardlu started on my questions....
[03-Feb-2011 12:08:57] <nyeates> Also, as a personal message from me - please keep in your minds the people of Egypt and the Arab world. I have some friends and family over there and we all hope that it works out for the best there.
[03-Feb-2011 12:09:25] <Sam-I-Am> nyeates: been keeping a close eye on it.
[03-Feb-2011 12:09:37] <Sam-I-Am> pbs newshour has awesome coverage
[03-Feb-2011 12:10:44] <nyeates> been watching aljazeera english
[03-Feb-2011 12:10:50] <Jane_Curry> Anyone done any deep digging into setting Hardware and OS Manufacturer and Model??
[03-Feb-2011 12:10:57] <Jane_Curry> I get quirky results
[03-Feb-2011 12:14:52] <rmatte> nyeates: where's the link to the beta?
[03-Feb-2011 12:14:59] <rmatte> oh nevermind
[03-Feb-2011 12:15:04] <rmatte> I see the link in the topic now
[03-Feb-2011 12:17:35] <kokey> whee, hopefully we'll get approval tomorrow to buy enterprise
[03-Feb-2011 12:17:45] <rmatte> cool
[03-Feb-2011 12:17:54] <kokey> i'm getting to a point of needing regular support
[03-Feb-2011 12:18:21] <rmatte> Jane_Curry: that whole system needs to have someone take a deep look at it, it hasn't worked right as long as I can remember
[03-Feb-2011 12:18:30] <kokey> or rather the list of questions have built up
[03-Feb-2011 12:18:43] <Sam-I-Am> kokey: i like the enterprise support crew