Open ivanich opened 5 years ago
Totally agree! All my pods consume around 3m or 10m or ~80m, except aws-service-operator exhausts CPU by consuming ~1000m !!
k top pod --all-namespaces
My temp solution k -n operators scale deployment aws-service-operator --replicas 0
that is terrible - consuming the CPU time of a whole node! Any idea in this regard?
One dirty workaround appears to be adding the CPU limits on the deployment - the helm chart has support for.
@christopherhein Thoughts on this issue?
JFYI, noticed issue with high cpu usage in the eks cluster for no reason, tried 1.0.0.alpha4 and self builds with alpine and ubuntu as base images. Deployed with helm charts btw