Unable to clear flapping

[WillemDH]

As we didn't receive emails for some services, I disabled flapping for them. Now it seems one of these services still has a flapping state and I seem unable to clear this state. I submitted 10+ passive check results to the service, so that the state change percentages is now 0%. But the service still has a flappign state.
So how can I remove the flapping state for this service?

I installed 5.2.7 today and rebooted and the flapping state is not going away.. Is this a known issue. Seems like that when flapping is disabled, services that already had the falpping state never get rid of it?

EDIT: I tried many things, such as re-enabling flapping and submitting a lots of OK results, but was not able to make the flapping state go away. So now I removed the service and copied it from a service which did not have this issue afetr which the problem was solved.

[tmcdonald]

I think this might be a bug, but we haven't (to my knowledge) heard this before. I'll have it tested and update the thread.

The alternative is to edit the retention.dat file, but that's messy.

[ischwartz]

Just letting you know we filed an internal feature request ID 8522:

We discussed internally that it would be a good idea to clear the flapping state when disabling flap detection

Keep an eye out on future Core releases.

[mcapra]

Since a feature request has been submit, is it ok if we close this thread?

[tmcdonald]

To add on, we weren't able to replicate 100% what you showed, but we did note that disabling flap detection leads to a situation where the state was stuck to whatever it was prior to disabling, but for us re-enabling did allow it to go away. I would double-check your service and see if you can force this behavior, as none of us were able to.

[BanditBBS]

Just to add on to this thread, i just ran into this scenario:

Service1 - Forgot to disable flap detection when added it
Active che3cks are disable
have a script that submits passive result with a specific message for an alert
script then submits passive result resetting back to OK


Well, after that scrip was called enough times it went into flapping state. I noticed and disabled flap detection. The flapping state never went away and i had to enable flap detection and submit many passive results to get it below threshold. Flapping reset and I then disabled detection again.

[lmiltchev]

@BanditBBS I tried recreating this issue again, but no luck. I disabled active checks on a Ping service, and made sure that the flap detection is NOT enabled. Next, I submitted bunch of passive check (CRITICAL & OK) until the service went to a flapping state. Then, I enabled flap detection. The "Service is flapping between states" message" disappeared, but the "State Change" didn't clear out (it was over 20%). I had to submit about 20 "OK" passive checks until the "State Change" percentage dropped below 5%.