World K2: Trays & Orders

Restart or replace?

You will learn: when a container restarts inside its pod, and when the whole pod is replaced.

Checked with Kubernetes 1.37 on 31 Oct 2026

The problem

Two trays had trouble this morning. One kept its name, and its RESTARTS column says 1. The other has a brand-new name. What is the difference?

Try it

A container crash. Press “Crash a container”, then run kubectl get pods. Watch the name and .

You are here: cluster kind-dev namespace default

dev-control-planeOpenManager

No trays here.

dev-workerOpen

shop·7z98n
shop·nq25c

dev-worker2Open

shop·blz2l

Time: +0s (sped up)

Watch

Want3written on the order for shop

See33 ready

Do All goodGaps fixed: 0

dev-workerOpen

shop·7z98n
shop·nq25c

dev-worker2Open

shop·blz2l

Now a whole tray goes. Delete one and watch. The manager makes a new tray with a new name. Kubernetes replaces pods; it does not repair them.

Try it

Restart in place, or a new tray? Sort what happens.

Drag a card into a box. Or tap the card, then tap the box.

Watch

You are here: cluster kind-dev namespace default

~/shop $ kubectl get podsNAME                   READY   STATUS    RESTARTS   AGEshop-747bf9cf8-7z98n   1/1     Running   0          22sshop-747bf9cf8-blz2l   1/1     Running   0          22sshop-747bf9cf8-nq25c   1/1     Running   0          22s~/shop $ kubectl rollout restart deployment/shopdeployment.apps/shop restarted~/shop $ kubectl rollout status deployment/shop --timeout=180sWaiting for deployment "shop" rollout to finish: 1 out of 3 new replicas have been updated...Waiting for deployment "shop" rollout to finish: 2 out of 3 new replicas have been updated...Waiting for deployment "shop" rollout to finish: 1 old replicas are pending termination...deployment "shop" successfully rolled out~/shop $ kubectl get podsNAME                    READY   STATUS        RESTARTS   AGEshop-747bf9cf8-7z98n    1/1     Terminating   0          43sshop-75c486fb7d-c7tkw   1/1     Running       0          14sshop-75c486fb7d-wv4zl   1/1     Running       0          7sshop-75c486fb7d-wwmpk   1/1     Running       0          21s

kubectl rollout restart replaces every tray, a few at a time. Same recipe, fresh trays. Compare the names before and after.

Check

You changed a setting the app reads only when it starts. How do all trays get it, with no downtime?

You did it!

You can now tell a container restart from a pod replacement.

Go deeper (optional)

Restarts wait longer each time: 10 s, 20 s, 40 s, up to 5 minutes. After 10 minutes without a crash, the wait resets. A new pod also gets a new IP address. That is why Services exist (World K3).

Real rep (optional): on a throwaway cluster, watch kubectl get pods -w during a rollout restart.