Container Logs and Output Streams

A pod is stuck in CrashLoopBackOff. You have maybe thirty seconds of useful context before the container restarts again. The answer is almost always in the logs, but only if you can pull them before the next restart wipes the slate.

"Manage and evaluate container output streams" is an explicit CKA competency within the Troubleshooting domain, which at 30% is the largest slice of the exam. The mechanics of how container logs work, where they live, and what kubectl logs can and cannot show you are the foundation of everything else in this module.