Multiple Nagios proccesses running.

[Andreas Ericsson]

Brian Murphy wrote:
> Andreas Ericsson <ae <at> op5.se> writes:
> 
> 
> 
>>Most definitely not. There's room for one method of processing checks, 
>>and one only (each method is quite complex and requires a fair amount of 
>>code). When (or if) the change is made, the old system of processing 
>>checks will be dropped entirely.
>>
>>
>>>Do I need to find a way to stop the results coming through in clumps?
>>>
>>
>>Not necessarily, no. You can probably get fairly good results by 
>>decreasing the service_reaper_frequency in nagios.cfg. Decrease it in 
>>steps of 2 until processes stop queuing up (that's when Nagios empties 
>>the pipe often enough for checks to get through).
>>
>>If it doesn't work, you'll need to split your nagios setup in smaller 
>>chunks or pay me a fat load of money to implement the thread-based or 
>>multiplexing checking model. Either way should work, really. ;)
>>
> 
> 
> 
> Currently trying 3 as the service reaper frequency, will try 1 on both the
> collectors and the central node and see how that goes.
> 

1 won't work (minimum scheduler interval is 1 second, and if it only 
checks for completed service checks it'll never have time to actually 
*do* any service checks), so Nagios won't allow it.

> Are there other config file settings that I should be changing to optimise the
> central servers processing of checks?
> 

Not sure, really. Grok the docs for the main config file. If you're in 
to C you could try removing the micro-sleep inside the reaper function 
(can't remember its name now) and replace them with sched_yield(), which 
basically has the same effect, but only if there are processes (or 
threads) with equal or greater priority that are pending. If you're 
running on a SMP box it'll also do funny (good) things to the task 
scheduler inside the kernel.

-- 
Andreas Ericsson                   andreas.ericsson at op5.se
OP5 AB                             www.op5.se
Lead Developer


-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
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