Ways and tweaks to make nagios more efficient. load average on monitoring host edging up.

[Rahul Nabar]

I set up my nagios system to monitor 256 odd nodes each with about 6
services (direct and NRPE). It is working fine but my load averages have
started edging upwards. Not critical yet but I wanted some tips to make
things more efficient and see if there are things I might have done
ineffeciently.

One of the points I identified is this: I am doing a ping and ssh check on
each server. This seems redundant. Is there a way to set it up so that:
Do a ssh check; if this succeds obviously ping is ok. If it fails do a ping
check and report on that.


How about the other way around too? I have a bunch of NRPE checks:
load_average, total-processes, scratch and home dir usage, pbs_mom,
ntp_time. If ssh fails then there is obviously no reason to try these other
checks right? But I think the monitoring_host wastes its cycles still trying
them (based on the "Last Check" time)

Any tips how I can achieve these effeciency tweaks? Or is there a problem in
my strategy? Any other performance tweaks so that I can squeeze every ounce
of Nagios performace?

Already I am using NRPE rather than check_by_sshh since I was told the
latter might be ineffecient for the monitoring host load usage.

-- 
Rahul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.monitoring-lists.org/archive/users/attachments/20090127/efd10757/attachment.html>
-------------- next part --------------
------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
-------------- next part --------------
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null