[kube-prometheus-stack] cant evaluate rules , promtheus getting oom after some queries failing and very less memory usage #10954
Replies: 3 comments
-
First, I think kube-prometheus discussion is a better place for this (https://github.com/prometheus-operator/kube-prometheus/discussions). Second, you can query by hand and check the results and try to track down the "why", in my experience this tends to happen when there are rests of an old Prometheus install (assuming you are Working on K8S and using the above-mentioned K8S setup). You also can disable the specified rules and check, also I don't think a rule evaluation error will trigger OOM, you can also check in Prometheus UI -> Status -> TSDB Stats to get an overview (or get a shell inside pod and run |
Beta Was this translation helpful? Give feedback.
-
It's unclear what's going on. We have fixed a bunch of on disk memory snapshots in the mean time, could be hepful if you could upgrade. |
Beta Was this translation helpful? Give feedback.
-
Also, if it was crashing while querying, the queries should appear in next startup logs, which I don't see here. |
Beta Was this translation helpful? Give feedback.
-
What did you do?
observed some weird behavior on prometheus .
What did you expect to see?
promtheus running smoothly without errors.
What did you see instead? Under which circumstances?
the pvc size is 50GB . and see the current data consumption of the disk when the oom happened.
prometheus resources
Environment
System information:
Darwin 21.4.0 x86_64
Prometheus version:
2.31.2
Prometheus configuration file:
Beta Was this translation helpful? Give feedback.
All reactions