[Nagiosplug-help] check_ifstatus returning false timeout?

[Thomas Guyot-Sionnest]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 10/01/08 07:24 AM, Philipp Geschke wrote:
> Hello List,
> 
> I am experiencing a strange behavior of check_ifstatus.
> 
> Since a couple of weeks, check_ifstatus is reporting a couple of alarm
> timeouts on 2 switches, which worked fine before.
> 
> First a assumed that there is something wrong with these switches, since
> I can't think of a reason for check_ifstatus to work that strangely.
> 
> But now I started monitoring this a little closer, and it turned out,
> that in fact check_ifoperstatus and a regular snmpwalk on these switches
> work at the same time that check_ifstatus is reporting a timeout.
> 
> Has anyone ever experienced this kind of strange behavior? I still could
> imagine that the nagios box' load is too high, but then other checks
> should have strange results too. Plus, at the time manually checking, I
> had a pretty low load:  load average: 0.05, 0.05, 0.12.
> 
> Here's a log of me manually triggering check_ifoperstatus,
> check_ifstatus and snmpwalk directly one after another, and this is no
>  coincidence:
> 
> nagios:/etc/nagios# /usr/local/nagios/libexec/check_ifoperstatus -k 46
> -H switch2 -C secret
> OK: Interface GigabitEthernet0/1/46 (index 46) is up.
> 
> nagios:/etc/nagios# /usr/local/nagios/libexec/check_ifstatus -C secret
> -H switch2
> ERROR: No snmp response from switch2 (alarm timeout)
> 
> nagios:/etc/nagios# snmpwalk -v1 -Of -c secret switch2
> mib-2.interfaces.ifTable.ifEntry.ifOperStatus
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.1
> = INTEGER: up(1)
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.2
> = INTEGER: up(1)
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.3
> = INTEGER: up(1)
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.4
> = INTEGER: down(2)
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.5
> = INTEGER: up(1)
> ------ SNIP ------
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.46
> = INTEGER: up(1)
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.47
> = INTEGER: down(2)
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.48
> = INTEGER: up(1)
> .iso.org.dod.internet.mgmt.mib-2.interfaces.ifTable.ifEntry.ifOperStatus.257
> = INTEGER: up(1)
> 
> I would appreciate any kind of hint here, even if it comes down to me
> just being too dumb for nagios ;-)

Network latency, packet drop and switch load can greatly influence the
time it takes to check the interfaces. Have you tried increasing the
timeout for check_ifoperstatus?

How much time it normally takes when run manually? Does it take
significantly longer sometimes (check with with at least 20-30 runs...)?

Thomas
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHhhn46dZ+Kt5BchYRAmUIAJ9AzX6Pqw4qRpP1Euah0KXtKPVazACguJSY
kIsa0XaUhxrEFqd0Kc2S+7E=
=8c0I
-----END PGP SIGNATURE-----