Hi again.
In doing some playing around with this, I realized that my original
description of the problem didn't well communicate my circumstances.
The Opsview monitor servers are different boxes than the ones whose RAID
controllers I want to monitor. From either of the monitor machines, call
them monA and monB, I can say on the command line 'check_megaraid -H
<target> -C <community>'. <target> is, of course, either of the two machines
of interest, call the targetC and targetD.
Fortunately, the check_megaraid script knows how to talk to the target
machines via SNMP, so as long as the appropriate firewall ports are
configured, I'm good to go.
>From your reply, Ton, I see that you are suggesting that we install the NRPE
daemon on each of monA and B along with whatever pieces of Nagios is
necessary to execute the check_megaraid remotely. However, because I
check_megaraid directly from the monitors, I have no need for remote
execution, and indeed, since the target machines are hosts for other VMs,
I'd rather not burden them with any other additional s/w overhead - it's
sufficient for my purposes that I can do it from the monitors.
It now appears that I can put check_megaraid in /usr/local/nagios/libexec
and build my service check directly on the monitor. The real question is,
what are the implications of doing so? Will Opsview delete the libexec
contents on upgrade possibly breaking my service check? If so, that may be a
problem. If not, all seems well.
Thanks for your help.
Sean
From: [email protected]
[mailto:[email protected]] On Behalf Of Ton Voon
Sent: Wednesday, December 22, 2010 06:05
To: Opsview Users
Subject: Re: [opsview-users] plugin not found in
/usr/local/nagios/libexec/nrpe_local
On 21 Dec 2010, at 20:53, Sean R. Kirkpatrick wrote:
I've got two instances of opsview running. Both machines have Dell Perc 4i
RAID controllers and I wanted to start monitoring their health using the
'check_megaraid' plugin. The instructions for adding a new plugin here
http://docs.opsview.org/doku.php?id=opsview3.10:unix_customise_agentseemed
straight forward and so I set out to make it work.
Unfortunately, it doesn't. When the plugin is placed in
/usr/local/nagios/libexec/nrpe_local with the appropriate config file entry
in /usr/local/nagios/etc/nrpe_local/override.cfg pointing to the script, the
script is never found - it does not show in the plugin dropdown list when
adding a new service check. Note that I have stopped and restarted
opsview-agent as well as nagios (opsview) itself.
So your issue is that the plugin does not get listed in the service check
edit pages in the drop down list?
This is not possible because this is a "local" plugin, so it doesn't make
sense for it to run from the Opsview master. You have to use check_nrpe to
execute it, so it will not show on the Opsview master (if the plugin was,
say, an OS/400 check, there's no way the Opsview master would be able to
represent that).
Your process of putting it in nrpe_local with an appropriate
command[check_megaraid] is the way to go. However, we don't do anything with
regards to distributing this new agent to your remote hosts.
Ton
_______________________________________________
Opsview-users mailing list
[email protected]
http://lists.opsview.org/lists/listinfo/opsview-users
