Unusual job duration alert with health checks

[ChrisJ]

Hi,

Is there a recommended way to handle this alert (Unusual job duration) when the jobs include healthchecks?
It will go over the 200% time whenever the healthcheck is scheduled.

I would have to increase the warning % too much to handle the healthcheck that the alert would be meaningless.
I never got the alert in v10, so something changed in v11.

Regards, Chris

[wishr]

Hi Christian,

Thank you for the feedback!

This is a new alarm of v11 that uses a relative calculation approach in comparison with the previously existing Max. allowed job duration alarm (that one is disabled by default) which does not rely on the historical data. The new alarm assesses the previous job sessions and compares their duration with the duration of the current sessions.

We'll check what can be improved in the alarm logic to handle the health checks.

Thanks

[Regnor]

I've stumbled over this today also with an executed health check, so it would be great if the runtime of those checks could be excluded.

[wishr]

Hi Max,

Sure, we'll try to improve that.