High CPU consumption
chelomontilla opened this issue · 3 comments
Hello,
I'm trying gatekeeper and found this great project to easily check config and violations.
I've deployed it in kubernetes using the helm chart and the CPU consumption is quite high with the latest relase version.
This is one pod with version 0.5.0
https://ibb.co/PYSHYHn
This is one pod with version 0.4.2
https://ibb.co/xfFfK6x
Let me know if I can collaborates somehow on solving the issue.
Thanks in advance
Hola @chelomontilla !
Nice finding, let me do some tests and I'll come back to you.
Just for my curiosity, is GPM v0.5.0 working as expected besides the high CPU usage? Do you see something in the logs? Can you try rising the log level to DEBUG and see if you find anything strange?
Thanks!
I can't replicate this locally. The CPU usage for v0.5.0 is under 1 milli core while GPM is idling:
$ kubectl top pod -n gatekeeper-system
NAME CPU(cores) MEMORY(bytes)
gatekeeper-audit-fbc6596f7-79pbj 12m 71Mi
gatekeeper-controller-manager-5f64c4d565-pnrjg 4m 51Mi
gatekeeper-controller-manager-5f64c4d565-sl4mz 5m 61Mi
gatekeeper-controller-manager-5f64c4d565-v4gkr 9m 62Mi
gatekeeper-policy-manager-7dfbd5cf55-2n2fk 1m 107Mi
$ kubectl describe pod -n gatekeeper-system gatekeeper-policy-manager-7dfbd5cf55-2n2fk| grep image
Normal Pulling 45m kubelet Pulling image "quay.io/sighup/gatekeeper-policy-manager:v0.5.0"
Normal Pulled 45m kubelet Successfully pulled image "quay.io/sighup/gatekeeper-policy-manager:v0.5.0" in 8.142802076s
Is there something special about your environment @chelomontilla that we should take into account?
Hello,
I've just checked and works perfectly. The high CPU consumption was due the memory requests and limits. Requests and limits for memory were configured with 128Mi and didn't see it was reaching limits. I've reconfigured the pod resources with 156Mi requests and 192Mi and the cpu is pretty low right now.
Thanks a lot