Page MenuHomePhabricator

Increase visibility of container/pod ressource exhaustion
Open, MediumPublic

Description

Currently we lack high level visibility of containers/pods that reach or exceed configured CPU/memory requests/limits as well as CPU throttling.

We should be able to detect that in a generic fashion to have issues like T266194 addressed more proactively.

I'll add this as a meta-task to track ideas/possible work regarding this.