Hosts are permanently up but shown as Flapping in availability

Troubleshooting
1

**CMK version:2.1.0p24
**OS version:RHEL 8

I am facing a issue withe the availability of some of my Hosts.
Round about 20 physical devices of my 600 monitored hosts (VMs, physical Hosts, Switches …) are shown as 100% flapping when I look at the “Availability - Status of Host XYZ” page. The Host it self is shown as UP and there are no changes of the Hoststate since the last updates of checkMK or the Host itself.
This problem I already facedin version 2.1.0p19 and when I look earlier in the Timerange of the availability, these Hosts also are shown as Flapping on all checkMK versions before (we just started have a look at the availability for KPIs).
These 20 Hosts are not of the same Devicetype or in the same subnet, the only thing they have in common ist that they are physical devices.
I have already performed a permanent ping from the cmk to those Hosts and there are no losses and only a few pings have a higher round trip time of 70ms than usually <1ms, but thats the same as other Hosts not shown as Flapping in availability.
Other things I hae already tried are change the “Host check command” from smart ping to normal ping, disabled flapp detection for those Hosts and disable flapp detection for all Hosts and changed tresholds/paremters for Ping and Flapdetection but nothing chaged.
The only thing that helped is to create a new Host in checkMK with the same IP, these are no more shown as Flapping int the availability page of the Host or in the Report. But it´s no option to monitor these Hosts with other Hostnames just to get rid of the Problem.

Has anyone else faced the same Problem and can share a fix?

2

Hi @chri,
that sounds like a bug. If there are no host state changes, the host should leave the flapping state pretty soon. Somehow maybe the core got mixed up? Are you using Nagios or the checkmk micro core?

If you set the host to “do not monitor” to have it’s current state removed from the state file, activate changes, then change it to “productive” again and activate changes again, I think the host history + perf data will still be there but the current state (including flapping or not flapping) should be re-calculated.

Gerd

3

Thanks for the quick respons,
we are running on checkmk micro core and i have set two of the hosts to “do not monitor” and back to production after a short while but it did not help. The Hoststate is still UP but in availability they are show as falpping exapt for the time they wre not monitored, there they are marked ad not available.
I have also deleted one of the Hosts and configured it new with the same hostname but seems that checkmk keeps the data of the old host and brings back to the new one with same name.

4

hm ok, I guess then it would be something worth passing to t29 dev using a support ticket :confused:

5

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.