High check latency with nagios

[Andreas Ericsson]

marino.simons at acerta.be wrote:
> In the meantime I discoverd that nagios waits 10 seconds when a check 
> times out (and repeats this 9 times),  so now I am looking where to change 
> this.

max_check_attempts in host objects (or their templates). You can safely 
set this to 5, but not lower than 3 unless you want flaky host-state 
detection.

> I did find another small error in our config and the number of 
> checks/minute is up from 50 to 280..  But I still find this rather low.. 
> Is there a workaround for the host check timeout?
> If I understand your first post correctly, nagios won't recover when my 
> network goes down?
> 
Yes it will, but not until it's soundly down and nagios knows it. This 
is to be 'cleverized', but is not so as of now (I know a good way to do 
it, but I don't have the time).

On the other hand, it's sort of silly to implement a monitoring solution 
if you know the network is riddled with problems. Nagios does its job 
best in a network that works, but where some points of it tends to fail 
or get overloaded from time to time.

> Thanks,
> Marino
> 
> 
> 
> 
> mshirley at ysi.com wrote:
> 
>>Also, just because you have 2 processors and smp
>>enabled in the kernel doesn't mean nagios was written with proper 
> 
> threading
> 
>>in mind (which defeats the purpose of dual processors).
>>
>>Not really. Nagios fork()'s. The child processes in turn execute the 
>>various plugins, which means that for every check to be run, there is 
>>actually always at least 3 threads to execute in parallell (main, child 
>>and plugin). Nagios doesn't have to bother with all this (although it 
>>could if one were to rewrite large parts of it in assembly), but rather 
>>lets the kernel handle it. The kernel is very good at it, so the loss is 
>>minimal compared to the developer-unfriendly asm options.
>>Point being; More CPU's (+ SMP support in kernel) = More power, no 
>>matter how the program was written or even compiled (although sending it 
>>once through strip is not a bad idea).
> 
> 
>>Mark Shirley
>>IT Helpdesk
>>YSI Incorporated
> 
> 

-- 
Andreas Ericsson
OP5 AB
+46 (0)733 709032
andreas.ericsson at op5.se


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE.
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
_______________________________________________
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