How to define critical limits for keepalive check?



I’m building a proof of concept with Sensu Go 5.2.1.

I can define keepalive-timeout and keepalive-interval in agent’s configuration file (agent.yml). But when the keepalive-timeout value is reached, the keepalive check gets it’s status changed to WARNING, and it stays there until the keepalive check is resolved.

But how can I get keepalive to get CRITICAL? Basically, I would like there to be a configuration flag keepalive-interval (as it is now), keepalive-warn-threshold (for example 120 seconds) and keepalive-crit-threshold (for example 240 seconds).

Am I getting something wrong or should I post a feature request into GitHub? :slight_smile:

Thanks in advance!


Hello @hezor,

I have confirmed the behavior you are seeing in the latest Sensu Go release. The keepalive event stays in warning even after 15 occurences. I’d expect that to continue happening.

Sensu 1.x has this concept, which you might be familiar with already:

I’d post a feature request against the Sensu Go project on GitHub. Include the above link to show that this was possible in Sensu 1.x, as I see users wanting to have keepalive events only be handled if they go critical. As a previous user, that was the model I went for. :-).


1 Like

from what i was told keepalive status is only ever 1 or 0. are you basing your critical threshold off of when you want to get alerted on the failed keepalive? if so, you could just filter the handler by occurrences.