[2009-02-05 10::01:25] mrayzenoss: Greetings
[2009-02-05 10::01:59] monrad: evening
[2009-02-05 10::02:27] mrayzenoss: Europe?
[2009-02-05 10::02:35] monrad: yes, Denmark
[2009-02-05 10::02:53] ** ju55i joined the chat room.
[2009-02-05 10::03:11] ** jmp242 joined the chat room.
[2009-02-05 10::03:49] jmp242: Hi zenoss devs
[2009-02-05 10::03:55] nemo_: oioi: same version, all ok here
[2009-02-05 10::04:34] oioi: rebooted ... not ok /(
[2009-02-05 10::07:23] ** zenChild joined the chat room.
[2009-02-05 10::07:28] nemo_: oioi, try zenperfsnmp run -v10
[2009-02-05 10::07:39] ** JraNil joined the chat room.
[2009-02-05 10::08:19] nemo_: oioi, rebooting unix is bad ...
[2009-02-05 10::08:29] oioi: hehe
[2009-02-05 10::08:42] gemineye: Hi mrayzenoss
[2009-02-05 10::08:49] jmp242: I'm wondering if anyone knows why thresholds don't seem to work for scripts like: http://forums.zenoss.com/viewtopic.php?t=8707&highlight=
[2009-02-05 10::08:49] nemo_: so just watch the whole snmp cycle oioi
[2009-02-05 10::09:04] adytum-bot: Title: Forums :: View topic - Threshold not generating an event (at forums.zenoss.com)
[2009-02-05 10::09:32] jmp242: I've successfully gotten graphing working per chet's suggestion here:
[2009-02-05 10::09:39] oioi: i get this
EBUG:zen.zenperfsnmp:unresponsive devices: [] nemo
[2009-02-05 10::09:41] JraNil: Hi every1, is there any list for recommended GSM Modem or Mobile Phone to work with SMS Alerting?
[2009-02-05 10::09:46] adytum-bot: Title: Forums :: View topic - Graphing number of status events (at forums.zenoss.com)
[2009-02-05 10::09:56] jmp242: but I don't get any events on exceeding the graph/datapoint thresholds
[2009-02-05 10::10:12] nemo_: well if zenoss isnt creating the RRD
[2009-02-05 10::11:50] oioi: i'm running as zenoss user
[2009-02-05 10::13:47] ** oioi left the chat room.
[2009-02-05 10::15:19] cgibbons: so what does your graph & threshold definitions look like, jmp?
[2009-02-05 10::16:21] C_Kode: Question. I have two JBoss servers and I'm using the JMX plugin. Both servers are running and both graphing, but one server isn't reporting ThreadCounts. It reports everything else though and I see ThreadCount in the JBoss console.
[2009-02-05 10::17:00] ** JraNil left the chat room.
[2009-02-05 10::20:11] ** npmccallum left the chat room.
[2009-02-05 10::20:41] gemineye: Is it possible to only display Warning and above level events on the dashboard?
[2009-02-05 10::20:45] jmp242: umm well, I basically did exactly what is in this post cgibbons
[2009-02-05 10::21:00] adytum-bot: Title: Forums :: View topic - Graphing number of status events (at forums.zenoss.com)
[2009-02-05 10::21:28] cgibbons: the post doesn't talk about creating any templates, datapoints, or thresholds... so you didn't do any of that, or?
[2009-02-05 10::21:37] jmp242: I didn't create any custom graph definitions
[2009-02-05 10::22:23] ** mrchippy joined the chat room.
[2009-02-05 10::22:48] jmp242: Then I created a threshold
[2009-02-05 10::22:59] cgibbons: c_kode: just checked with one of the devs that worked on the JMX stuff a lot. he's wondering if you can see that value using jconsole remotely.
[2009-02-05 10::23:11] jmp242: called overlicense, with licenses_checkedout highlighted
[2009-02-05 10::23:28] ** Vog-work left the chat room.
[2009-02-05 10::23:34] mrayzenoss: gemineye: Which portlet?
[2009-02-05 10::23:38] jmp242: left defaults and set enabled to true
[2009-02-05 10::24:19] ** Vog-work joined the chat room.
[2009-02-05 10::24:47] cgibbons: how about the severity and escalate count?
[2009-02-05 10::25:04] gemineye: Any/all portlets. For us it's only important if an event an event is really Error or higher but Warning adds nicely to the color pallet.
[2009-02-05 10::25:29] jmp242: severity is Warning and escalate count is 0
[2009-02-05 10::27:44] mrayzenoss: gemineye: You set the Dashboard Priority Threshold right?
[2009-02-05 10::27:51] ** mikl_ joined the chat room.
[2009-02-05 10::30:17] ** |cico| left the chat room.
[2009-02-05 10::34:48] jmp242: is this something I should be trying to do in the python script, perhaps with zensendevent?
[2009-02-05 10::35:57] mrayzenoss: gemineye: the answer I got from support is it's only done in code and should be exposed as an option, assuming that you really want "Dashboard Severity Threshold"
[2009-02-05 10::36:01] cgibbons: perhaps, i think the current script just doesn't apply thresholds after collecting the data. it looks like in zenoss processing the thresholds is the job for each collector daemon, and since this one is a bit out of that scope it never happens.
[2009-02-05 10::37:50] jmp242: ok, off to learn some python then, thanks
[2009-02-05 10::40:22] mrayzenoss: gemineye: opened http://dev.zenoss.org/trac/ticket/4429
[2009-02-05 10::41:33] cgibbons: jmp, I wonder if you'd be better off not running it from cron
[2009-02-05 10::42:14] jmp242: yea, the whole cron thing was chet's idea
[2009-02-05 10::42:22] cgibbons: that way the zencommand daemon would get the value, put it in the graph, and then apply thresholds
[2009-02-05 10::42:33] jmp242: I was wondering if it was a performance issue to use command datasources?
[2009-02-05 10::42:38] cgibbons: chet knows a whole lot more about zenoss than I do so there might be a good reason for that, but it doesn't really fit the normal data collection cycle
[2009-02-05 10::43:12] ** mikl left the chat room.
[2009-02-05 10::43:17] jmp242: At this point, I'm not really sure how to capture the value output from the script (I don't know python, currently on my to do list)
[2009-02-05 10::43:58] cgibbons: lesee, that might be pretty easy, sec
[2009-02-05 10::44:01] jmp242: It doesn't look like the script is ever putting the count into a value
[2009-02-05 10::44:34] cgibbons: the row[0] expression at line 27 does it
[2009-02-05 10::45:16] C_Kode: cgibbons: yep, I can see them in the web-console and in the jconsole from the Zenoss server
[2009-02-05 10::45:26] cgibbons: huh okay
[2009-02-05 10::45:56] jmp242: Hmm, it looks like it's incrementing the rrd file there? So what I'd want to do is instead in that loop increment a variable?
[2009-02-05 10::47:49] ** klinstifen joined the chat room.
[2009-02-05 10::48:01] cgibbons: so you have the admin guide .pdf handy?
[2009-02-05 10::48:30] C_Kode: moi or jmp?
[2009-02-05 10::48:36] cgibbons: jmp
[2009-02-05 10::48:48] jmp242: umm, I have the website open?
[2009-02-05 10::49:01] cgibbons: sent another ping to the jmx guru, c_kode. one thing he asked earlier was if you saw any errors when running zenjmxor in the logs.
[2009-02-05 10::49:20] C_Kode: Looking.
[2009-02-05 10::50:21] cgibbons: that chapter walks you through creating a script or using a script and what output format it wants (which is the nagios plugin format). you could take that script chet built and just have it output the string as needed instead and then have zencommand run it.
[2009-02-05 10::50:25] C_Kode: I see why it doesn't show up now.
[2009-02-05 10::50:25] cgibbons: as an option
[2009-02-05 10::50:29] C_Kode: rrd error not a simple integer: '309.0' Devices/asct4/ZenJMX Thread Count_ThreadCount
[2009-02-05 10::50:38] cgibbons: k let me forward that to him
[2009-02-05 10::50:39] C_Kode: There are 309 threads currently on that server.
[2009-02-05 10::51:08] jmp242: Yes, I do know how to do command datasources. I'll look into that as well. Thanks
[2009-02-05 10::52:10] cgibbons: that script would change slightly to take a single device argument, which would the need to be passed into the event query... and then you'd only expect one row.
[2009-02-05 10::52:47] ** gemineye left the chat room.
[2009-02-05 10::53:02] ** gemineye joined the chat room.
[2009-02-05 10::54:13] jmp242: I'm thinking I just don't understand this enough - wouldn't I get many rows for the different existing events?
[2009-02-05 10::55:06] cgibbons: that query will summarize all the events for each device separately, just like it does now
[2009-02-05 10::56:31] C_Kode: looking
[2009-02-05 11::00:19] cgibbons: yep
[2009-02-05 11::00:24] gemineye: mrayzenoss: Sorry, someone popped into my office and wouldn't leave.
[2009-02-05 11::00:30] cgibbons: are the two jboss server versions the same?
[2009-02-05 11::00:41] C_Kode: Yep, there clustered together
[2009-02-05 11::00:47] cgibbons: even more freaky
[2009-02-05 11::01:07] C_Kode: Yep.
[2009-02-05 11::02:12] cgibbons: can you verify the datapoint type for that threadcount datapoint?
[2009-02-05 11::02:54] C_Kode: Gauge
[2009-02-05 11::03:18] gemineye: mrayzenoss: I don't understand the ticket #4429. Shouldn't the Dashboard Priority Threshold do what I'm asking? If not what is it for?
[2009-02-05 11::03:40] mrayzenoss: gemineye: there's both a severity and a priority on the event
[2009-02-05 11::04:15] cgibbons: which template are you looking @, C_kode? i'll look @ the same here.
[2009-02-05 11::04:39] C_Kode: Since RRD seems to not like the fact that it's returning a decimal. (312.0) you sould figure it would have the other server too
[2009-02-05 11::04:55] cgibbons: yeah
[2009-02-05 11::05:10] C_Kode: I'm looking at /ZenJMX/ZenJMX Thread Count/ThreadCount
[2009-02-05 11::08:05] cgibbons: so what's odd here is that the same data point here is a COUNTER type
[2009-02-05 11::08:12] ** charlieS_ left the chat room.
[2009-02-05 11::08:16] ** charlieS joined the chat room.
[2009-02-05 11::08:22] cgibbons: what version of ZenJMX are you using? (settings -> zenpacks)- I don't want to look @ the wrong one
[2009-02-05 11::08:45] gemineye: Oh, ok. All my events currently have a priority of -1 (None). How do most people avoid having annoying things like bad OID info events appearing on the dashboard?
[2009-02-05 11::09:11] C_Kode: 3.0
[2009-02-05 11::09:35] cgibbons: okay, same as here, and we have it as COUNTER
[2009-02-05 11::09:52] C_Kode: Alright, lets do some changing and see whats happening
[2009-02-05 11::10:12] cgibbons: was that the first server may have originally created its RRD file with the counter type, it got changed and then then 2nd server created the RRD file with the other datatype
[2009-02-05 11::10:42] C_Kode: So I need to delete the RRD files?
[2009-02-05 11::10:56] cgibbons: let's just try for the bad device
[2009-02-05 11::14:18] C_Kode: ok, waiting for the next checkl
[2009-02-05 11::17:03] ** jmp242 left the chat room.
[2009-02-05 11::17:55] C_Kode: Hmm.. I didn't see the error in the log when it ran the check, but ThreadCount is zero on the graph
[2009-02-05 11::18:18] gemineye: Does anyone monitor network devices that are connected using MPLS? We can't get the lines to be drawn between sites because of this.
[2009-02-05 11::19:17] C_Kode: Yep it doesn't appear as nan anymore. Now it's flat zero
[2009-02-05 11::19:30] cgibbons: and you have 0 threads, no usage!
[2009-02-05 11::19:35] C_Kode: heh
[2009-02-05 11::20:21] mrayzenoss: gemineye: back to your question about priorities vs. severities, priority is assigned on Devices, Groups, Systems and Locations, so you can indicate that a certain things are more important than others
[2009-02-05 11::25:35] ** cico joined the chat room.
[2009-02-05 11::29:29] mrayzenoss: Thanks for everyone who showed up, the transcript will be online soon.
[2009-02-05 11::30:03] ** You changed the topic to "Zenoss is Open Source Application, Systems & Network Monitoring".
[2009-02-05 10::01:59] monrad: evening
[2009-02-05 10::02:27] mrayzenoss: Europe?
[2009-02-05 10::02:35] monrad: yes, Denmark
[2009-02-05 10::02:53] ** ju55i joined the chat room.
[2009-02-05 10::03:11] ** jmp242 joined the chat room.
[2009-02-05 10::03:49] jmp242: Hi zenoss devs
[2009-02-05 10::03:55] nemo_: oioi: same version, all ok here
[2009-02-05 10::04:34] oioi: rebooted ... not ok /(
[2009-02-05 10::07:23] ** zenChild joined the chat room.
[2009-02-05 10::07:28] nemo_: oioi, try zenperfsnmp run -v10
[2009-02-05 10::07:39] ** JraNil joined the chat room.
[2009-02-05 10::08:19] nemo_: oioi, rebooting unix is bad ...
[2009-02-05 10::08:29] oioi: hehe
[2009-02-05 10::08:42] gemineye: Hi mrayzenoss
[2009-02-05 10::08:49] jmp242: I'm wondering if anyone knows why thresholds don't seem to work for scripts like: http://forums.zenoss.com/viewtopic.php?t=8707&highlight=
[2009-02-05 10::08:49] nemo_: so just watch the whole snmp cycle oioi
[2009-02-05 10::09:04] adytum-bot: Title: Forums :: View topic - Threshold not generating an event (at forums.zenoss.com)
[2009-02-05 10::09:32] jmp242: I've successfully gotten graphing working per chet's suggestion here:
[2009-02-05 10::09:39] oioi: i get this
EBUG:zen.zenperfsnmp:unresponsive devices: [] nemo[2009-02-05 10::09:41] JraNil: Hi every1, is there any list for recommended GSM Modem or Mobile Phone to work with SMS Alerting?
[2009-02-05 10::09:46] adytum-bot: Title: Forums :: View topic - Graphing number of status events (at forums.zenoss.com)
[2009-02-05 10::09:56] jmp242: but I don't get any events on exceeding the graph/datapoint thresholds
[2009-02-05 10::10:12] nemo_: well if zenoss isnt creating the RRD
[2009-02-05 10::11:50] oioi: i'm running as zenoss user
[2009-02-05 10::13:47] ** oioi left the chat room.
[2009-02-05 10::15:19] cgibbons: so what does your graph & threshold definitions look like, jmp?
[2009-02-05 10::16:21] C_Kode: Question. I have two JBoss servers and I'm using the JMX plugin. Both servers are running and both graphing, but one server isn't reporting ThreadCounts. It reports everything else though and I see ThreadCount in the JBoss console.
[2009-02-05 10::17:00] ** JraNil left the chat room.
[2009-02-05 10::20:11] ** npmccallum left the chat room.
[2009-02-05 10::20:41] gemineye: Is it possible to only display Warning and above level events on the dashboard?
[2009-02-05 10::20:45] jmp242: umm well, I basically did exactly what is in this post cgibbons
[2009-02-05 10::21:00] adytum-bot: Title: Forums :: View topic - Graphing number of status events (at forums.zenoss.com)
[2009-02-05 10::21:28] cgibbons: the post doesn't talk about creating any templates, datapoints, or thresholds... so you didn't do any of that, or?
[2009-02-05 10::21:37] jmp242: I didn't create any custom graph definitions
[2009-02-05 10::22:23] ** mrchippy joined the chat room.
[2009-02-05 10::22:48] jmp242: Then I created a threshold
[2009-02-05 10::22:59] cgibbons: c_kode: just checked with one of the devs that worked on the JMX stuff a lot. he's wondering if you can see that value using jconsole remotely.
[2009-02-05 10::23:11] jmp242: called overlicense, with licenses_checkedout highlighted
[2009-02-05 10::23:28] ** Vog-work left the chat room.
[2009-02-05 10::23:34] mrayzenoss: gemineye: Which portlet?
[2009-02-05 10::23:38] jmp242: left defaults and set enabled to true
[2009-02-05 10::24:19] ** Vog-work joined the chat room.
[2009-02-05 10::24:47] cgibbons: how about the severity and escalate count?
[2009-02-05 10::25:04] gemineye: Any/all portlets. For us it's only important if an event an event is really Error or higher but Warning adds nicely to the color pallet.
[2009-02-05 10::25:29] jmp242: severity is Warning and escalate count is 0
[2009-02-05 10::27:44] mrayzenoss: gemineye: You set the Dashboard Priority Threshold right?
[2009-02-05 10::27:51] ** mikl_ joined the chat room.
[2009-02-05 10::30:17] ** |cico| left the chat room.
[2009-02-05 10::34:48] jmp242: is this something I should be trying to do in the python script, perhaps with zensendevent?
[2009-02-05 10::35:57] mrayzenoss: gemineye: the answer I got from support is it's only done in code and should be exposed as an option, assuming that you really want "Dashboard Severity Threshold"
[2009-02-05 10::36:01] cgibbons: perhaps, i think the current script just doesn't apply thresholds after collecting the data. it looks like in zenoss processing the thresholds is the job for each collector daemon, and since this one is a bit out of that scope it never happens.
[2009-02-05 10::37:50] jmp242: ok, off to learn some python then, thanks
[2009-02-05 10::40:22] mrayzenoss: gemineye: opened http://dev.zenoss.org/trac/ticket/4429
[2009-02-05 10::41:33] cgibbons: jmp, I wonder if you'd be better off not running it from cron
[2009-02-05 10::42:14] jmp242: yea, the whole cron thing was chet's idea
[2009-02-05 10::42:22] cgibbons: that way the zencommand daemon would get the value, put it in the graph, and then apply thresholds
[2009-02-05 10::42:33] jmp242: I was wondering if it was a performance issue to use command datasources?
[2009-02-05 10::42:38] cgibbons: chet knows a whole lot more about zenoss than I do so there might be a good reason for that, but it doesn't really fit the normal data collection cycle
[2009-02-05 10::43:12] ** mikl left the chat room.
[2009-02-05 10::43:17] jmp242: At this point, I'm not really sure how to capture the value output from the script (I don't know python, currently on my to do list)
[2009-02-05 10::43:58] cgibbons: lesee, that might be pretty easy, sec
[2009-02-05 10::44:01] jmp242: It doesn't look like the script is ever putting the count into a value
[2009-02-05 10::44:34] cgibbons: the row[0] expression at line 27 does it
[2009-02-05 10::45:16] C_Kode: cgibbons: yep, I can see them in the web-console and in the jconsole from the Zenoss server
[2009-02-05 10::45:26] cgibbons: huh okay
[2009-02-05 10::45:56] jmp242: Hmm, it looks like it's incrementing the rrd file there? So what I'd want to do is instead in that loop increment a variable?
[2009-02-05 10::47:49] ** klinstifen joined the chat room.
[2009-02-05 10::48:01] cgibbons: so you have the admin guide .pdf handy?
[2009-02-05 10::48:30] C_Kode: moi or jmp?
[2009-02-05 10::48:36] cgibbons: jmp
[2009-02-05 10::48:48] jmp242: umm, I have the website open?
[2009-02-05 10::49:01] cgibbons: sent another ping to the jmx guru, c_kode. one thing he asked earlier was if you saw any errors when running zenjmxor in the logs.
[2009-02-05 10::49:20] C_Kode: Looking.
[2009-02-05 10::50:21] cgibbons: that chapter walks you through creating a script or using a script and what output format it wants (which is the nagios plugin format). you could take that script chet built and just have it output the string as needed instead and then have zencommand run it.
[2009-02-05 10::50:25] C_Kode: I see why it doesn't show up now.
[2009-02-05 10::50:25] cgibbons: as an option
[2009-02-05 10::50:29] C_Kode: rrd error not a simple integer: '309.0' Devices/asct4/ZenJMX Thread Count_ThreadCount
[2009-02-05 10::50:38] cgibbons: k let me forward that to him
[2009-02-05 10::50:39] C_Kode: There are 309 threads currently on that server.
[2009-02-05 10::51:08] jmp242: Yes, I do know how to do command datasources. I'll look into that as well. Thanks
[2009-02-05 10::52:10] cgibbons: that script would change slightly to take a single device argument, which would the need to be passed into the event query... and then you'd only expect one row.
[2009-02-05 10::52:47] ** gemineye left the chat room.
[2009-02-05 10::53:02] ** gemineye joined the chat room.
[2009-02-05 10::54:13] jmp242: I'm thinking I just don't understand this enough - wouldn't I get many rows for the different existing events?
[2009-02-05 10::55:06] cgibbons: that query will summarize all the events for each device separately, just like it does now
[2009-02-05 10::56:31] C_Kode: looking
[2009-02-05 11::00:19] cgibbons: yep
[2009-02-05 11::00:24] gemineye: mrayzenoss: Sorry, someone popped into my office and wouldn't leave.
[2009-02-05 11::00:30] cgibbons: are the two jboss server versions the same?
[2009-02-05 11::00:41] C_Kode: Yep, there clustered together
[2009-02-05 11::00:47] cgibbons: even more freaky
[2009-02-05 11::01:07] C_Kode: Yep.
[2009-02-05 11::02:12] cgibbons: can you verify the datapoint type for that threadcount datapoint?
[2009-02-05 11::02:54] C_Kode: Gauge
[2009-02-05 11::03:18] gemineye: mrayzenoss: I don't understand the ticket #4429. Shouldn't the Dashboard Priority Threshold do what I'm asking? If not what is it for?
[2009-02-05 11::03:40] mrayzenoss: gemineye: there's both a severity and a priority on the event
[2009-02-05 11::04:15] cgibbons: which template are you looking @, C_kode? i'll look @ the same here.
[2009-02-05 11::04:39] C_Kode: Since RRD seems to not like the fact that it's returning a decimal. (312.0) you sould figure it would have the other server too
[2009-02-05 11::04:55] cgibbons: yeah
[2009-02-05 11::05:10] C_Kode: I'm looking at /ZenJMX/ZenJMX Thread Count/ThreadCount
[2009-02-05 11::08:05] cgibbons: so what's odd here is that the same data point here is a COUNTER type
[2009-02-05 11::08:12] ** charlieS_ left the chat room.
[2009-02-05 11::08:16] ** charlieS joined the chat room.
[2009-02-05 11::08:22] cgibbons: what version of ZenJMX are you using? (settings -> zenpacks)- I don't want to look @ the wrong one
[2009-02-05 11::08:45] gemineye: Oh, ok. All my events currently have a priority of -1 (None). How do most people avoid having annoying things like bad OID info events appearing on the dashboard?
[2009-02-05 11::09:11] C_Kode: 3.0
[2009-02-05 11::09:35] cgibbons: okay, same as here, and we have it as COUNTER
[2009-02-05 11::09:52] C_Kode: Alright, lets do some changing and see whats happening
[2009-02-05 11::10:12] cgibbons: was that the first server may have originally created its RRD file with the counter type, it got changed and then then 2nd server created the RRD file with the other datatype
[2009-02-05 11::10:42] C_Kode: So I need to delete the RRD files?
[2009-02-05 11::10:56] cgibbons: let's just try for the bad device
[2009-02-05 11::14:18] C_Kode: ok, waiting for the next checkl
[2009-02-05 11::17:03] ** jmp242 left the chat room.
[2009-02-05 11::17:55] C_Kode: Hmm.. I didn't see the error in the log when it ran the check, but ThreadCount is zero on the graph
[2009-02-05 11::18:18] gemineye: Does anyone monitor network devices that are connected using MPLS? We can't get the lines to be drawn between sites because of this.
[2009-02-05 11::19:17] C_Kode: Yep it doesn't appear as nan anymore. Now it's flat zero
[2009-02-05 11::19:30] cgibbons: and you have 0 threads, no usage!
[2009-02-05 11::19:35] C_Kode: heh
[2009-02-05 11::20:21] mrayzenoss: gemineye: back to your question about priorities vs. severities, priority is assigned on Devices, Groups, Systems and Locations, so you can indicate that a certain things are more important than others
[2009-02-05 11::25:35] ** cico joined the chat room.
[2009-02-05 11::29:29] mrayzenoss: Thanks for everyone who showed up, the transcript will be online soon.
[2009-02-05 11::30:03] ** You changed the topic to "Zenoss is Open Source Application, Systems & Network Monitoring".
