nagios 3.0b3 abnormal cpu load ?

[Anthony Montibello]

Hi List,

First I think this is more appropiate fpr the Developer list so I hope Ethan
has been watching this thread.

If your observation is true,
should Nagios have an internal default hostcheck command  or some code to
deactivate hostchecks when no hostcheck command is entered?

Tony (author of NC_Net)


On 9/10/07, SCHAER Frederic <frederic.schaer at cea.fr> wrote:
>
>  I finally found a solution :
>
> either deactivate the host checks or… define a check_command for the hosts
> : with host_checks_enabled and without a check command, nagios eats all of
> my CPU, with a command defined, everything is back to normal…
>
>
>
> I wish I saw I had no check_command defined in my host template !
>
>
>
> Cheers
>
>
>   ------------------------------
>
> *From:* SCHAER Frederic
> *Sent:* Monday, September 10, 2007 5:56 PM
> *To:* SCHAER Frederic; nagios-users at lists.sourceforge.net
> *Subject:* RE: [Nagios-users] nagios 3.0b3 abnormal cpu load ?
>
>
>
> /me again…
>
>
>
> It seems there is a problem with host checks on my box that causes host
> checks to be run continuously on the same host(s) – well it seems that there
> is indeed no host check run because one is « always » executing ( ? no, it's
> not ! ).
>
>
>
> logs show this for a debug run of just a few seconds :
>
> # grep "A check of this host is already being executed"
> /var/log/nagios/nagios.debug | wc -l
>
>    1709
>
>
>
> de-activating host checks (execute_host_checks=0)in the nagios.cfg file
> frees the CPU, at the cost of host checks… it's good to know, isn't it ? ;)
>
> Is this really usefull anyway if I define a check_ping service on all
> hosts ?
>
>
>
> Cheers
>
>
>   ------------------------------
>
> *From:* nagios-users-bounces at lists.sourceforge.net [mailto:
> nagios-users-bounces at lists.sourceforge.net] *On Behalf Of *SCHAER Frederic
> *Sent:* Monday, September 10, 2007 12:25 PM
> *To:* nagios-users at lists.sourceforge.net
> *Subject:* Re: [Nagios-users] nagios 3.0b3 abnormal cpu load ?
>
>
>
> Hi again,
>
>
>
> There is a difference between the nagios I compiled on the 2.4 kernel /
> RHEL3 box on the one hand, and the 2.6 one on Ubuntu on the other hand :
>
> when stracing the nagios process, I can see it's calling the nanosleep()
> function – and it then sleeps -, whereas on the 2.4 box I can't see such a
> call : I just see this kind of thing (and lots of these) :
>
>
>
> gettimeofday({1189419621, 161574}, NULL) = 0
>
> time([1189419621])                      = 1189419621
>
> time([1189419621])                      = 1189419621
>
> gettimeofday({1189419621, 183742}, NULL) = 0
>
> gettimeofday({1189419621, 183780}, NULL) = 0
>
> gettimeofday({1189419621, 183814}, NULL) = 0
>
> time([1189419621])                      = 1189419621
>
> gettimeofday({1189419621, 184172}, NULL) = 0
>
> gettimeofday({1189419621, 184326}, NULL) = 0
>
> time([1189419621])                      = 1189419621
>
> time([1189419621])                      = 1189419621
>
> gettimeofday({1189419621, 184734}, NULL) = 0
>
> gettimeofday({1189419621, 184861}, NULL) = 0
>
> gettimeofday({1189419621, 184984}, NULL) = 0
>
> time([1189419621])                      = 1189419621
>
> gettimeofday({1189419621, 185537}, NULL) = 0
>
> gettimeofday({1189419621, 185734}, NULL) = 0
>
>
>
> Bad thing is that I tried to configure with and without the
> –enable-nanosleep or –disable-nanosleep switches, but this apparently has no
> effect :'(.
>
> So it seems nagios is sleeping all the time, eating the CPU… I'm still
> trying to see if I can find a way to compile with a "working sleep".
>
>
>
> Fred
>
>
>
>
>   ------------------------------
>
> *From:* nagios-users-bounces at lists.sourceforge.net [mailto:
> nagios-users-bounces at lists.sourceforge.net] *On Behalf Of *SCHAER Frederic
> *Sent:* Friday, September 07, 2007 5:51 PM
> *To:* nagios-users at lists.sourceforge.net
> *Subject:* [Nagios-users] nagios 3.0b3 abnormal cpu load ?
>
>
>
> Hi,
>
>
>
> I'm trying to setup a nagios 3.0b3 server, and I configured it with only
> one host, one check (ping).
>
> I can successfully start nagios, but I have concerns about the server cpu
> load : that's around 99% cpu usage all the time, on a P4 3GHz machine…
>
>
>
> Is there something I could do to decrese this load ? I tried installing a
> nagios 2.9 server using the same configuration, and the load caused by
> nagios is just … 0.
>
> I guess this is a bug in the beta 3 ?
>
>
>
> Increasing the sleep time in nagios.cfg does not help at all even if I put
> sleep_time=2500 (2500 seconds !)
>
>
>
> Any suggestions would be welcome J
>
>
>
> P.S : I compiled nagios on a 2.4.21 linux kernel, using gcc 3.2.3, don't
> know if this could help. I don't see many usefull things with maximum debug
> verbosity…
>
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> 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/20070910/8f75787a/attachment.html>
-------------- next part --------------
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
-------------- next part --------------
_______________________________________________
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