没有运行的作业，但是cpu居高不下，如何能够让作业释放资源？ about cube-studio HOT 1 CLOSED

删除旧的训练任务
kubectl get workflow -n pipeline | awk '{print $1}' | xargs kubectl delete workflow --force --grace-period=0 -n pipeline
kubectl get tfjob -n pipeline | awk '{print $1}' | xargs kubectl delete tfjob --force --grace-period=0 -n pipeline
kubectl get pytorchjob -n pipeline | awk '{print $1}' | xargs kubectl delete pytorchjob --force --grace-period=0 -n pipeline
kubectl get mpijob -n pipeline | awk '{print $1}' | xargs kubectl delete mpijob --force --grace-period=0 -n pipeline
kubectl get deployment -n pipeline | awk '{print $1}' | xargs kubectl delete deployment --force --grace-period=0 -n pipeline

清理旧的jupyter
kubectl get pod -n jupyter | awk '{print $1}' | xargs kubectl delete pod --force --grace-period=0 -n jupyter

清理历史service
kubectl get deployment -n service | awk '{print $1}' | xargs kubectl delete deployment --force --grace-period=0 -n service