A Kubernetes rollout is stuck halfway. Old pods are running, new ones are not coming up. What is going on?
What they are really testing: Whether you understand that a stuck rollout is usually Kubernetes protecting you, and can enumerate the reasons a new pod cannot schedule or become ready.
A real interview question
A Kubernetes rollout is stuck halfway. Old pods are running, new ones are not coming up. What is going on?
What most people say
drag me
“I would delete the stuck pods so Kubernetes recreates them.”
The recreated pod hits the identical condition, so it changes nothing except adding churn. It also skips the diagnosis entirely: the reason the pod cannot become ready is the answer to the question.
The follow-ups they ask next
The events say Insufficient cpu. What are your options?
Lower the request if it was set too high, scale the node group, or check whether a pod disruption budget or leftover pods hold capacity. Also check whether requests grew in this release.
Why might maxSurge and maxUnavailable matter here?
They control how much room the rollout has. maxSurge 0 means it must remove an old pod before adding a new one, which can deadlock when capacity is tight or a disruption budget prevents removal.
What the interviewer is listening for
- Frames the stall as a safety feature
- Distinguishes Pending from not-ready
- Notes there is no user impact so no need to rush
What sinks the answer
- Deletes pods to force it
- Cannot enumerate scheduling failures
- Panics despite old pods serving
If you genuinely do not know
Say this instead of freezing. Reasoning out loud from what you do know beats silence every single time, and a good interviewer is listening for exactly that.
“This is [Kubernetes protecting me: old pods still serve until new ones are ready]. I check [Pending means it never scheduled: resources, selectors, taints, volumes], [running but not ready means image pull, missing config or a failing readiness probe]. Since [there is no user impact], I [roll back if the cause is not obvious quickly] and investigate calmly.”
Keep going with troubleshooting
Junior
A developer says the build passes locally but fails in CI. How do you debug that?
Junior
A production server is failing writes. Disk is 100% full. Walk me through what you do.
Junior
At midnight everything started failing with TLS errors. What happened and what do you do?
Mid
A pod is in CrashLoopBackOff after a deploy. Walk me through your debugging, command by command.
Mid
The deploy reports success, pods are Running, but users get 502s. Where do you look?
Mid
A Kubernetes node goes NotReady and pods are stuck Terminating. What is happening and what do you do?
Knowing the answer is not the same as recalling it under pressure
Sign in to send the questions you fumble to spaced recall, so they come back right before you would forget them, and learn the concepts behind them with hands-on labs.
Start free