Bug#292473: [Pkg-nagios-devel] Bug#292473: acknowledged by
developer (Bug#292473: fixed in nagios 2:1.3-cvs.20050402-1)
Tue, 19 Apr 2005 17:16:14 +0200
Le mardi 19 avril 2005 =E0 09:21 -0400, sean finney a =E9crit :
> hi cyril,
> On Tue, Apr 19, 2005 at 04:03:43PM +0300, Cyril Bouthors wrote:
> > As I told you, the server is constantly loaded by all those cronjobs
> > and daemons, the load average is between 1 and 4.
> > Nagios should not fail in this situation.
> this is true, but i'm wondering what makes you situation so different
> from other people's (my nagios server has had loads like this and
> not exhibited this behavior).
BTW, there are many ways to improve performance of checks in Nagios.
Using other checks, preferably more optimized, like check_fping instead
of check_ping, activating the embedded Perl whenever possible, ... are
all good ways to improve all that.
For example, I used to monitor a lot of machines some years ago with
Netsaint, and monitoring Oracle can be really heavy if you use the
official Oracle client to do that. The main performance issue was the
RAM needs of all the check plugins, so playing with their scheduling so
that they don't start all together helped a lot.
> > max_concurrent_checks=3D5
> could you see if scaling this down, and/or check_interval for
> your services in services.cfg has any effect?
Jerome Warnier <email@example.com>