CMK version: 2.3.0p29
OS version: Debian 12.10
SNMP-Alarms obviously wrong, but also for agentbased Clients
For some time now, we have been experiencing sporadic mass alarms for accessibility via SNMP.
The curious thing: SNMP devices are demonstrably O.K. But agent-based systems are also alerted.
Here are two screenshots.
Yes - in your error message you see “mgmt_…”. This device has a management interface configured with SNMP.
Please remove this configuration and then the error should be gone.
I would advice to create the management interface as a separate host object inside CheckMK for monitoring.
Damn 
That’s right!
Unfortunately, the only thing that remains is the phenomenon that we sporadically receive dozens or hundreds of SNMP alarms - like the one shown in the screenshot.
And we have no clue what the cause could be.
Is SNMP monitoring working on these devices?
The error only means that there is an SNMP timeout but it is not visible why.
Yes, that’s the crazy thing: all SNMP-based devices work perfectly. The messages are simply wrong.
Once or twice a day this message appears, as if Checkmk itself had a problem.
There can only be a problem on the CheckMK server if there are not enough resources available to do all the checks.
I think the reason is actually a lack of resources. A second instance was also running on the hardware (test for Openshift monitoring).
I have uninstalled this (omd rm XXXX) and am now observing the phenomenon.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed. Contact an admin if you think this should be re-opened.
