Failure Handling and Restart Policies

LeaderWorkerSet provides configurable failure handling for pod groups when failures occur.

Restart Policies

RecreateGroupOnPodRestart (Default)

When any pod in a group fails, the entire group is recreated. This ensures all pods start fresh together.

spec:
  leaderWorkerTemplate:
    restartPolicy: RecreateGroupOnPodRestart

Use case: Tightly coupled applications (distributed influence)

None

Only the failed pod is restarted. Other pods in the group are not affected.

spec:
  leaderWorkerTemplate:
    restartPolicy: None

RecreateGroupAfterStart (Experimental)

When any pod in a group fails, the entire group is recreated if and only if there are no pods currently pending. This feature is currently experimental on versions 0.8+.

apiVersion: leaderworkerset.x-k8s.io/v1
kind: LeaderWorkerSet
metadata:
  name: leaderworkerset-sample
  annotations:
    leaderworkerset.sigs.k8s.io/experimental-recreate-group-after-start: true

Node Failure Handling

With RecreateGroupOnPodRestart (default): When a node fails, the entire group is recreated on healthy nodes.

With None: Only pods on the failed node are rescheduled. Other pods continue running.