giantswarm / roadmap

Giant Swarm Product Roadmap
https://github.com/orgs/giantswarm/projects/273
Apache License 2.0
3 stars 0 forks source link

SLO alert label update #2579

Open TheoBrigitte opened 1 year ago

TheoBrigitte commented 1 year ago

Goal

We need to be able to change label on SLO alert. e.g. a component ownership changes and team label need to be updated in the alert.

Problem

Changing SLO alert seems not possible currently as this cause prometheus rules evaluation failure today after SLO alert label got changed.

Details

Changes:

This lead to the following image

If you change any of those labels, while keeping other values the same, then the expressions in those rules will start to return more than one series (one with the old label, one with the new label), causing the recording rules to fail until the old label ages out.

More details :

TheoBrigitte commented 1 year ago

Solutions

TheoBrigitte commented 1 year ago

This happened again today, someone changed the team label on a SLO alert which broke the alert.

We need to make sure people do not do this again: