Nagios Support Forum

> -----Original Message-----
> From: [email protected] [mailto:nagios-devel-
> [email protected]] On Behalf Of loren jan wilson
> Sent: Sunday, August 13, 2006 2:43 PM
> To: [email protected]
> Subject: [Nagios-devel] ocsp slows nagios a great deal
>=20
> I'm in the process of trying to set up a distributed nagios
> environment monitoring about 9,000 services on 2,500 hosts. I'm using
> Sunfire V210 servers running Solaris 10.
>=20
> I've found that the distributed servers which monitor the active
> services can run about 1700 checks every 5 minutes if ocsp isn't
> enabled, but once I enable obsess_over_services, the number of active
> checks I can do goes WAY down. Is this known behavior? It seems like
> it might be a bug or something...why would it be part of the design to
> lose so much performance per server when implementing distributed
> monitoring?
>=20
> Here's a breakdown:
>=20
> - ocsp disabled: 1800 checks / 5 min.
>=20
> - ocsp command set to /bin/true: 1200 checks / 5 min.
>=20
> - ocsp command set to a perl program that forks, then pipes output to
> send_nsca: 800 checks / 5 min.
>=20
> - ocsp command set to a shell program that pipes output to send_nsca
> (this is the "submit_check_result_via_nsca" script that came with
> the nagios distribution): 500 checks / 5 min.

FWIW I'm able to perform at least 1300 checks / 5 min on a distributed
server reporting to two central hosts each accepting ~3500 results via
nsca every 5 minutes (from multiple distributed nagios installations).
Each distributed server sends the results to the two central machines
via nsca resulting in ~2600 submissions every 5 minutes.

This is under RHAS3 with really no special tuning done and using the
documented OCSP procedure for distributed checks. Have you done any
testing to determine where the time penalty is originating?

--
Marc





This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]