Flapping trigger

[TuXator]

Hi!

I have a host that has some agent items, one simple item (ssh) and a status item configured.

If the agent is unavailable (not running), the ssh check flaps between OK and unknown (see screenshot).

Probably as a consequence, the status trigger also flaps between OK and PROBLEM. However, the logs on the host show that the ssh daemon is correctly contacted every 90 seconds (as configured in the ssh item).

Any hints why the ssh simple check doesn't work correctly if the agent is not running?

[Justleen]

We are experiencing exactly the same issue. Can any one shed any light on this?
I have tried changing the value store to "as is" and even if then we get an "Unknown" status on ssh.
The "Unknown" usually lasts only seconds (10 on avg.) but it is enough to start triggers (and mess up the statistics in the reports/top 10 triggers)