RES: Urgent: nagios + apan memory chewing up

[Marcio Queiroz]

I have 200 services using apan and i have the same problem, but it only
happens when i increse the number_of_concurrent_checks for more than 30. The
problem is that if i limit in 30, my check latency goes for something like 5
minutes. If the solution is to take off apan, i would like to know. Any
ideias ?

[]s

	Marcio Q.  Dall Agnol
	phone : 55-21-3873-5826
	Fax      : 55-21-3873-5845
	marcioqueiroz at vicom.com.br
	VICOM
	Net Serviços de Comunicação S/A
	Visite nossas páginas:
	Redes Corporativas: <http://www.vicom.com.br>
	Internet Banda Larga: <http://www.virtua.com.br>
	TV por Assinatura : <http://www.net.tv.br>
	Institucional: http://www.netservicos.com



> ----- Mensagem original -----
> De:		Ian Davidson [SMTP:mav at juniper.net]
> Enviada em:		quarta-feira, 10 de setembro de 2003 09:59
> Para:		alikhalidi at excite.com; chet at rcn.com;
> nagios-users at lists.sourceforge.net
> Cc:		marcioqueiroz at vicom.com.br
> Assunto:		RE: [Nagios-users] Urgent: nagios + apan memory
> chewing up
> 
> What was the final conclusion to all of this? I've just hit the same
> problem of swap-space being chewed up to the point of distraction but
> cannot pinpoint from this thread what the exact -fix- is. Is it really
> the perl stuff that's causing it to barf?
> 
> Cheers,
> 
> Ian
> 
> 
> 
> ________________________________
> 
> From: nagios-users-admin at lists.sourceforge.net
> [mailto:nagios-users-admin at lists.sourceforge.net] On Behalf Of ali
> Sent: 16 August 2003 08:30
> To: chet at rcn.com; alikhalidi at excite.com;
> nagios-users at lists.sourceforge.net
> Cc: marcioqueiroz at vicom.com.br
> 
> 
> 
> Thanks for the reply, and I realy appriciate your input.
> I have my nagios installed from an RPM, version 1.1
> I've read somewhere in the posts about this recommendation (compile
> without embedded per and perl cache), but after I've checed the RPM spec
> built file, it seems that the person who built the RPM has disabled
> these two options for RedHat 8 and 9. so I am running a version with
> these two options off (any other suggestions), but I might have mis-read
> the sepc file!.
> 
> So you're telling me that nagios (v 1.1) has this core problem, even
> without Embeded Perl and Perl Cache, or is it this factor that you
> eleminated to get around the problem, please clearify the issue
> pertaining to the nagios plugings (especially check_snmp)?
> Has the disabling of these two options solved the problem completely in
> your case (because if so, I can drop ! the RPMS and go for source)?
> 
> thanks,
> 
> 
> 
> 
> 
> --- On Thu 08/14, Chet Luther < chet at rcn.com > wrote:
> 
> 
> 	From: Chet Luther [mailto: chet at rcn.com]
> 	To: alikhalidi at excite.com, nagios-users at lists.sourceforge.net
> 	Cc: marcioqueiroz at vicom.com.br
> 	Date: Thu, 14 Aug 2003 11:00:13 -0400
> 	Subject: Re: [Nagios-users] Urgent: nagios + apan memory chewing
> up
> 	
> 	Ali,
> 	Recompile Nagios without embedded perl or perl caching support.
> I had
> 	this same problem, and it was due to global namespace pollution
> in the
> 	embedded perl cache. The real solution to this problem would be
> rewriting
> 	all the perl plugins/extensions in a mod_perl safe way, but I
> don't see any
> 	initiative to get that done.
> 	
> 	Hope this helps,
> 	
> 	Chet Luther
> 	chet at rcn.com
> 	
> 	----- Original Message ----- 
> 	From: ali
> 	To: nagios-users at lists.sourceforge.net
> 	Sent: Thursday, August 14, 2003 5:40 AM
> 	Subject: [Nagios-users] Urgent: nagios + apan memory chewing up
> 	
> 	! Greetings,
> 	
> 	I am experiencing a strange behaiviour for nagios and apan.
> 	I'de run nagios for a long time, using the standard plugins
> 	(check_host_alive for 50 services), and my system
> 	(CPU 1GHz, RAM 512, swap 512) was very stable.
> 	not untill I've installed apan (155 services) that the server
> starts chewing
> 	memory so fast that the system becomes cloged and the only way
> to regain
> 	control for the system is to REBOOT.
> 	I've read and practiced a lot with the tunning paramers of
> nagios, and
> 	reached a level where these parameters
> 	max_concurrent_checks (for concurrency)
> 	service_reaper_freq
> 	size of rrd files (to prevent caching of large databse files)
> 	added the -m to snmp_get (based on a recommendation of one of
> the users) to
> 	disable the loading of the mib with each call
> 	
> 	but all with no help. I can't put my hands on the problem.the
> system starts
> 	memory usage in almost a linear fashion, untill the free section
> of
> 	buffers/cache st! arts comming down, and the systems goes to
> swap till it
> 	exhast all system Memory.
> 	it is only the difference between the concurrency factor
> 	(max_concurrent_checks) that slows down the rate at which memory
> is lost.
> 	
> 	its a production system, and I really appriciate help
> 	
> 	thanks,
> 	
> 	
> 
> 
> 	
> 
> ________________________________
> 
> Join Excite! - http://www.excite.com
> The most personalized portal on the Web! 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> 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


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
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