Page 1 of 1

Collecting Perf Data

Posted: Fri Feb 22, 2013 7:09 pm
by ssawin
I am new to NagiosXI and graphing data in general. I am adding some custom checks...this one for example. https://www.monitoringexchange.org/inve ... monitoring
The checks work fine and report the data but I want the data to be graphed. What dictates what can and cannot be graphed?

I imported my check plugins and used the core config manager to create a custom command and then attached that to a hostgroup and stuck my switches in that hostgroup but none of these have the graph icon next to them. Graphing is currently working great in other parts for example ping. How can I add graphing to my switch load and cpu usage checks?

Thanks

1.What version of Nagios XI are you using? - 2012R1.6
2.Linux Distribution and version? - VMware Image
3.32 or 64bit? - 64Bit
4.VMware Image or Manual Install of XI? - VMware Image
5.Are there specials configurations on your system, ie; is Gnome installed? Are you using a proxy? Are you using SSL? - Nope

Re: Collecting Perf Data

Posted: Mon Feb 25, 2013 8:30 am
by scottwilkerson
ssawin wrote:What dictates what can and cannot be graphed?
The plugin must return performance data
http://nagiosplug.sourceforge.net/devel ... tml#AEN201

Re: Collecting Perf Data

Posted: Fri Mar 29, 2013 10:31 am
by emarmonti
I have a question related to this. I'm also using Nagios XI.

I'm having trouble finding an answer to this question from google searches, so now I'm asking here. Does anyone know if there is a limit on the number of performance lables you can provide in the output of one service check?

I created a service check that pipes out close to 60 data collection points that I'm trying to graph. When viewing this service in Nagis XI, and clicking the Performance Graphs link, I see all of the graph labels for this service, but only the first label shows a graph. Additionally, none of them seem to actually be recording data.

Per http://nagiosplug.sourceforge.net/devel ... tml#AEN201, I think I'm recording the values correctly (space delimited, not using single quotes because there is no space in the label names). Though it does say "Be aware of a limitation in the amount of data that NRPE returns to Nagios" - I just don't know if the same applies to Nagios XI?

Anyway, it's not recording data. Since there are close to 60 labels passed for performance data, I'm wondering if I'm pushing the limits of something - but I don't know (I've only been using nagios with graphing for about 6 months)

Here is the output from this check:

./check_zfs_cpuInfo -h brickstor1
CPU Metrics Info | bawrite=0 bread=0 bwrite=0 canch=220 class=0 cpu_load_intr=3 cpu_nsec_idle=2818629022890590184978975293440 cpu_nsec_intr=215268255715311 cpu_nsec_kernel=748712892621741 cpu_nsec_user=184775523730530 cpu_ticks_idle=3049044782 cpu_ticks_kernel=74871280 cpu_ticks_user=18477540 cpu_ticks_wait=0 cpumigrate=1386531149 crtime=3810648352565 idlethread=8662276205 intr=215315458140622 intrblk=7638494 intrthread=5703860403 intrunpin=0 inv_swtch=19487914 iowait=0 lread=0 lwrite=0 mdmint=0 modload=252 modunload=78 msg=0 mutex_adenters=4109981600 namei=303170800 nthreads=5725080 outch=11378649 phread=59744819 phwrite=1866 procovf=0 pswitch=17984822596 rawch=105796 rcvint=0 readch=14387859724676 rw_rdfails=14716768 rw_wrfails=334706 sema=0 snaptime=2789241291975491 syscall=31613256673 sysexec=1853016 sysfork=1871038 sysread=14329598111 sysvfork=97890 syswrite=14036189992 trap=2099081421 ufsdirblk=0 ufsiget=0 ufsinopage=0 ufsipage=0 wait_ticks_io=0 writech=13907015461393 xcalls=3281355013 xmtint=0




Thanks

Re: Collecting Perf Data

Posted: Sat Mar 30, 2013 12:15 pm
by scottwilkerson
Is it possible that the RRD was created with just one DS the first time it was run? and the plugin has been changed after that?

Re: Collecting Perf Data

Posted: Tue Apr 02, 2013 8:48 am
by emarmonti
scottwilkerson wrote:Is it possible that the RRD was created with just one DS the first time it was run? and the plugin has been changed after that?
That's a good question. I'll have to do some research / review some documentation and see if I can figure out how to flush the RRD (or the equivalent) for this service only.

Something else I thought of, this check is spitting out 64 bit integers - maybe Nagios XI / the RRD can't handle 64 bit numbers?



EJ

Re: Collecting Perf Data

Posted: Tue Apr 02, 2013 10:34 am
by lmiltchev
I believe RRDTool can handle both, 32- and 64-bit.

http://oss.oetiker.ch/rrdtool/tut/rrdtutorial.en.html

Re: Collecting Perf Data

Posted: Tue Apr 02, 2013 2:21 pm
by emarmonti
lmiltchev wrote:I believe RRDTool can handle both, 32- and 64-bit.

http://oss.oetiker.ch/rrdtool/tut/rrdtutorial.en.html

Thanks for the link.
I don't know exactly what went wrong, but I simply renamed the service check and now it's collecting data properly.

Re: Collecting Perf Data

Posted: Tue Apr 02, 2013 2:28 pm
by slansing
Ah excellent, thank you for letting us know what solved the problem. Maybe it simply needed to grip onto the configuration again.