spark-operator
spark-operator copied to clipboard
spark executor pods not releasing memory when idle
I am running spark structure streaming on k8s. when the job is idle the memory uses of the executor pods is not coming down. is the running pod release memory or it reserves all the memory assign to it?