Skip to content

ProgressDeadlineExceeded

The Deployment controller marked the rollout as failed because no new pod became available for progressDeadlineSeconds (default 600). Kubernetes stops here; it doesn’t roll back.

  1. Find the new ReplicaSet (highest revision): kubectl get rs -l app=web --sort-by=.metadata.creationTimestamp.
  2. Look at its pods: kubectl get pods -l app=web and pick the newest. Their state names the real problem: ImagePullBackOff, CrashLoopBackOff, Pending, or Running but not Ready.
  3. If no new pods exist at all, the ReplicaSet couldn’t create them: kubectl describe rs <new-rs> shows quota or Pod Security rejections.
Terminal window
kubectl rollout undo deployment/web
kubectl rollout status deployment/web

With GitOps, revert the commit instead, or the controller will reapply the broken version.

Then follow the error page for whatever state the new pods were in.

Read next: Deployments and rollouts.