MidTroubleshooting

Your CI pipeline has been green for months. This morning every build fails at the deploy step with an authentication error against the cloud provider. Nobody touched the pipeline. Walk me through it.

What they are really testing: They are testing whether you can stabilise a blocked team quickly and then attack the structural cause, which is usually long lived static keys that finally expired or got rotated.

A real interview question

Your CI pipeline has been green for months. This morning every build fails at the deploy step with an authentication error against the cloud provider. Nobody touched the pipeline. Walk me through it.

What most people say

drag me

I would generate a new access key, paste it into the CI secrets, and the build goes green again. Problem solved.

It fixes today and guarantees the same outage in ninety days, and it shows no curiosity about why the credential died in the first place.

The follow-ups they ask next

  • How does OIDC federation actually stop this happening again?

    Explain there is no stored secret at all: the CI provider signs a short lived token, the cloud trusts the issuer and the specific repo, and the session lasts minutes.

  • What if the key was deactivated deliberately by security and you just re-minted it?

    Admit that would be the real mistake, and say you always confirm with the owner before restoring a credential that something else killed on purpose.

What the interviewer is listening for

  • Reads the exact error code and reasons from it
  • Unblocks the team first with a time box, then hunts cause
  • Checks whether the key was killed intentionally before reviving it
  • Proposes OIDC to remove the whole class of failure

What sinks the answer

  • Rotates the key and calls it done
  • Cannot distinguish expired from revoked from unauthorised
  • Stores the new key in plaintext in the repo or a shared doc
  • Never asks who or what disabled the credential

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.

I have not owned a CI credential outage, but I would reason like this: confirm how wide the failure is, read the exact auth error, unblock safely, then find what expired and remove the long lived key entirely.

Keep going with troubleshooting

All 336 cloud engineer questions

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