NagiosXI 5.2.3 CentOS CHECK_ESX3-0.5.PL time out.

[Unigarant]

Hi,

After a recent upgrade of NagiosXI to 5.2.3 our VDI check for available machines/hosts started generating lots of timouts.
This is done using the CHECK_ESX3-0.5.PL. (check_esx3_dc).

Nothing has changed on the virtual center side of things.
We have had this issue before, I think it involved the version of the perl libraries on our machines...

We tried upping the -t parameter to 30 but this did not help. IT seems the check just does not complete and generates a timeout after 1 minute.
Is there a way to get this check to run smoothly again and prevent the time out?

It is very important to get this check up and running again. Right now we are getting allerts about every 10 minutes and we probably won't notice if something is really wrong with our VDI enviroment...

Please let me know if you need more info to fix this problem!

[rkennedy]

Does it work if you run it with -t 60 or even higher?

[Unigarant]

Does it work if you run it with -t 60 or even higher?

Tested it with -t 61:

This is what I get when the check does not complete (this is just part of it, the actuall text is many times longer).

Code: Select all

... >vm-11549</obj></returnval><returnval><obj type="VirtualMachine">vm-11858</obj></returnval><returnval><obj type="VirtualMachine">vm-11636</obj></returnval><returnval><obj type="VirtualMachine">vm-11027</obj></returnval><returnval><obj type="VirtualMachine">vm-11859</obj></returnval><returnval><obj type="VirtualMachine">vm-11026</obj></returnval><returnval> ...

When I run the check again and it works it usually gives me the output within 1 second:

Code: Select all

CHECK_ESX3-0.5.PL OK - 327/401 VMs up, 11/11 Hosts up, overall status=gray, no config issues | vmcount=327units;; hostcount=11units;; config_issues=0;;

I have not seen the first error before, only after setting the timeout to 61 and waiting for a long time did I get this error. I don't know if it is because of the timeout value or because I don't usually wait half an hour for it to complete...

[rkennedy]

To clarify - when you ran it the second time, did you run it with -t 61 as well?

[Unigarant]

To clarify - when you ran it the second time, did you run it with -t 61 as well?

Yes I did. Exact same options, but it finishes in a few seconds instead of timing out/giving the above output.

[rkennedy]

Thanks for the clarification. Are any of the VMware host servers under a heavy load?

If not, I'd guess it to be a dependency. What OS are you running?

[Unigarant]

Thanks for the clarification. Are any of the VMware host servers under a heavy load?

If not, I'd guess it to be a dependency. What OS are you running?

Some of the servers might get as high as 85%, I have not seen higher.
We are running on CentOS, 6 I think. If you need the exact version please tell me how to check (ver?)?

We have had this issue before and that had something to do with a Perl module being updated. I think we downgraded it and where fine after that...

[bwallace]

Use this cmd to get the exact version:

cat /etc/centos-release

[Unigarant]

Use this cmd to get the exact version:

cat /etc/centos-release

6.7 final.

[Box293]

I would try replacing the plugin with the check_vmware_api.pl:
http://git.op5.org/gitweb?p=system-addo ... ads/master

They renamed it to this and this is a much more recent version.