Hi,
We have created one severity level to handle resolved alert issues:
Following are the details:
1. Resolved alerts are not differentiable between *what is actually
condition not matching with firing* and *metric not found* or
*condition not met *
- Keeping the metric absence check on every rule is cumbersome
2. Resolved alerts are having last firing result , however it is known
proemtheus design pattern.
- how to deal with it, can we add expression as part of
description to evaluate as part of annotations, how much load will it be on
prometheus and how to write the expressions ?
So, having above 2 issues, we have added 'CLEARED' as an another severity
and disabled send_revolved flag. Will it be the right way to do it or any
other suggestions to do similar
and also do you have any case studies to go over large deployments E2E life
cycle with proemtheus and alertings, remote storage.
Please suggest, it will help us and the community to go stronger with
better understanding.
Thanks in advance.
Regards,
Rajesh
--
You received this message because you are subscribed to the Google Groups
"Prometheus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/prometheus-users/CAEyhnp%2Bhzh2bsccHzjsdV-rK2Z_CYUXY2td0UbS214Tj2HB5vQ%40mail.gmail.com.
