The Job That Never Finishes

The situation

A nightly data-import Job in the kubelings namespace never finishes. The downstream step waits for kubectl wait --for=condition=complete job/data-import and hangs forever. The pod looks "Running" indefinitely.

A Job is complete only when its container exits 0. This one runs sleep infinity — it never exits, so the Job stays active forever.

Your task

Make data-import run to completion (a Job named data-import in kubelings with status.succeeded ≥ 1 and condition Complete=True).

kubectl -n kubelings get job data-import -o yaml | less
kubectl -n kubelings logs job/data-import

Job pod templates are immutable — to change the command you must delete and recreate the Job.

Hint

kubectl -n kubelings delete job data-import
# recreate with a command that does the work and EXITS 0, e.g.:
#   command: ["sh","-c","echo importing...; sleep 2; echo done"]

See solution.md.

Solution

Root cause

The Job's container ran sleep infinity. A Job only reports Complete when its pod's container exits 0 — a process that never exits keeps the Job Active forever.

Fix

Job pod templates are immutable, so delete and recreate with a command that finishes:

kubectl -n kubelings delete job data-import

kubectl -n kubelings apply -f - <<'EOF'
apiVersion: batch/v1
kind: Job
metadata:
  name: data-import
spec:
  backoffLimit: 4
  template:
    spec:
      restartPolicy: OnFailure
      containers:
        - name: importer
          image: ghcr.io/iximiuz/labs/busybox:latest
          command: ["sh","-c","echo importing...; sleep 2; echo done"]
EOF

Verify

kubectl -n kubelings wait --for=condition=complete job/data-import --timeout=60s
kubectl -n kubelings get job data-import

Course progress

The Job That Never Finishes

The situation

Your task

Root cause

Fix

Verify