Deleting old pods using CronJob on OpenShift

Whenever I do a cluster install and after running a cluster for a while, the “Completed” pod references start to mount up in the GUI. Here’s a quick and dirty way to clean up old pods.

apiVersion: v1
kind: ServiceAccount
metadata:
  name: pod-cleaner
  namespace: openshift
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: pod-cleaner-role
rules:
- apiGroups: [""]
  resources: ["pods"]
  verbs: ["list", "delete"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: pod-cleaner-binding
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: pod-cleaner-role
subjects:
- kind: ServiceAccount
  name: pod-cleaner
  namespace: openshift
---
apiVersion: batch/v1
kind: CronJob
metadata:
  name: cleanup-completed-pods
  namespace: openshift
spec:
  schedule: "0 2 * * *"   # daily at 02:00
  jobTemplate:
    spec:
      template:
        spec:
          serviceAccountName: pod-cleaner
          restartPolicy: Never
          containers:
          - name: pod-cleaner
            image: registry.redhat.io/openshift4/ose-cli:latest
            env:
              # Inline your target namespaces here (one per line)
              - name: NAMESPACES
                value: |
                  openshift
                  openshift-marketplace
                  openshift-operator-lifecycle-manager
                  openshift-kube-scheduler
                  openshift-etcd
                  openshift-kube-api-server
                  openshift-storage
            command:
              - /bin/sh
              - -c
              - |
                set -eu
                echo "Starting pod cleanup run at $(date -Is)"
                printf "%s\n" "$NAMESPACES" | while read -r ns; do
                  [ -z "$ns" ] && continue
                  echo "Cleaning namespace: $ns"
                  oc delete pods -n "$ns" --field-selector=status.phase=Succeeded || true
                  oc delete pods -n "$ns" --field-selector=status.phase=Failed || true
                done
                echo "Cleanup complete at $(date -Is)"
            resources:
              requests:
                cpu: 50m
                memory: 64Mi
              limits:
                cpu: 200m
                memory: 256Mi