Max number of services that can be monitored ?

[Hendrik Baecker]

Hi all,

I would suggest, that Alex should enable the perf-data, to see if there 
is one host or special service that screws up the latency.
We have one Server with double proceccor Xeon 1,0Ghz, 2 Gig of Ram and 
since Nagios 2.0b3 no latency above 5 seconds. Except there are many 
hosts done in our network, but this works as designed.
Together we are monitoring more than 4000 Services on nearly 1000 Hosts. 
EMBEDDED Perl ist disabled and in relation to the other posts here we 
have many perl scripts running as service checks. I don't belive that it 
helps much when you cancel the check_ntp check.

1. Enable the perf-data. (Tip: Say Nagios it should write both, host and 
service perf data in one file. So you can see how Nagios works and when 
the latency screws up)
2. Perhaps, even if the memory is not eaten up by nagios, think about to 
give your server more RAM.
3. Try to figure out, if your system has sometimes wa-cycles. If every 
process on your system is waiting for IO, then the ram is NOT eaten up 
and swapping isn't used to, but even no performance.

Please tell us, what service_check interval you have. If your nagios 
checks nearly every minute it may be hard for: a) your monitoring host, 
b) the (local) network, c) the (possibly existing) wan network.

Are your 60 hosts normaly up and running or do they often falls down?
Everytime when nagios is executing a host check, all other service 
checks are beeing "stopped", because a host check gets a higher check 
priority then a service check. So if  the scheduler has enough to do for 
service checks and there are some host checks comming between the other 
service checks are executed later then they normaly were scheduled.

Hope that gives some ideas.

Hendrik

misc at viceconsulting.co.nz schrieb:

>Hi Rob,
>
>  
>
>>You haven't sent in any information on what your server is actually
>>doing, so we can't offer much help there.  Amount of ram seems low, how
>>much RAM is free when you're running Nagios?
>>
>>What standard service checks are you running for each server?
>>
>>Are you running any of the Perl plugins (and having to launch Perl each
>>time you do a service check?)
>>    
>>
>
>The Nagios server is a dedicated Nagios server, it is not doing anything
>beyond Nagios.  It is monitoring 60 hosts.  Each host has 16 services
>configured on it.  Of those 16, 14 are run on the local host via
>check_nrpe, the remaining 2 are run on the Nagios server and are check_ssh
>and check_ntp (check_ntp is a Perl script.)
>
>The nagios is also self monitoring itself via check_disk, check_load,
>check_procs and a couple of shell scripts.
>
>Just out of interest, if you have NOT enabled embedded Perl, then am I
>correct in saying that Nagios will INVOKE a Perl script in the exact same
>way as it will invoke a C plugin, such as check_nrpe ?  ie it will do a
>fork() and run a given program, be it Perl, Shell, C etc.
>
>I have about 160MB of RAM free, and that amount has been free for several
>days, swap hasn't been touched.
>
>Should I remove the NTP check, is it really going to be such an issue
>given its only 1 of 16 services is using Perl?
>
>  
>
>>>Is what's happening with my Nagios server normal, or have I made a blatant
>>>configuration error that is slowing everything down?
>>>
>>>
>>>      
>>>
>>Lack of information provided, so can't provide any clues.
>>
>>    
>>
>>>I'm runinng Nagios 2.0b4 (embedded Perl DISABLED) on redhat es 3 update 5.
>>>
>>>
>>>      
>>>
>>As with the service checks.  With the Embeded Perl interpereter
>>disabled, if you run any Perl based checks, you will have to launch Perl
>>each and every time.
>>    
>>
>
>With embedded Perl enabled, I was getting memory leaks which was
>eventually crashing the Nagios server so I disabled it.
>
>If you can provide any further advice that would be much appreciated.
>
>Cheers,
>Alex.
>
>
>-------------------------------------------------------
>This SF.Net email is sponsored by:
>Power Architecture Resource Center: Free content, downloads, discussions,
>and more. http://solutions.newsforge.com/ibmarch.tmpl
>_______________________________________________
>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
>
>
>  
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.monitoring-lists.org/archive/users/attachments/20051011/577202a1/attachment.html>