I guess you mean the nagios_cache that I should disable? 

Actually I'm running 10 gmetad threads.
Is there a rule of thumb for the number of those threads? 


Regards,
Alex




From:   Vladimir Vuksan <vli...@veus.hr>
To:     Alexander Karner/Germany/IBM@IBMDE, 
Cc:     ganglia-general@lists.sourceforge.net
Date:   21.05.2012 16:53
Subject:        Re: [Ganglia-general] Ganglia <--> Icinga/Nagios 
Integration



I have observed the same situation. There is a race condition if you are 
caching metrics and the way Nagios/Icinga schedules checks where you may 
end up with an empty cache. Way to resolve this is to either

1. Turn off cache - you will need to increase number of gmetad threads
2. Warm up the cache e.g.

https://github.com/ganglia/ganglia-web/tree/master/nagios

contains a warm up script you can run every e.g. 10 seconds that populates 

the cache.

Vladimir


On Mon, 21 May 2012, Alexander Karner wrote:

> Hi!
> 
> I'm actually facing an interesting behaviour of the nagios scripts in 
the Ganglia web UI:
> (Using the latest available versions of Ganglia and Ganglia Web on AIX)
> 
> 
> 1) Sometimes a critical state is returned - allthough the last beacon 
was received within the threshold:
> 
> $ ./check_heartbeat host=host1 threshold=50
> CRITICAL Last beacon received 0 days, 0:00:11
> 
> --> Running this command from the command line repeatedly may or may not 
return this kind of information or an OK statement. It does
> not matter if the last beacon was received within the last 1, 3, 15 or 
"whatever" seconds.
> 
> 2) I also installed Icinga to monitor the health of the Ganglia daemons.
> In the service details I see up to now for each single host either the 
OK statement - or I get a "UNKNOWN Hostname info not
> available. Likely invalid hostname"
> This message ususally disappears after some or several additionally 
attempts to get the data but may come back later.
> 
> If I try to verify the state from the comand line I receive in most 
cases the "OK" statement and in a few cases the above shown
> "CRITICAL" value.
> 
> Does anyone know how to bring this in a more consistent state?
> 
> Mit freundlichen Grüßen / Kind regards
> 
> Alexander Karner
> 
> 
>------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Ganglia-general mailing list
Ganglia-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-general

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Ganglia-general mailing list
Ganglia-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-general

Reply via email to