REST API, mission 3 of 7
CrashLoopBackOff
intermediate, about 10 minutes, in the Kubernetes Simulator
Start mission
Opens once you pass Read at start.
The briefing
The last deploy of orders dropped something. The pods start, stop, start again — and each time Kubernetes waits a little longer before trying: CrashLoopBackOff.
A crash loop is the container exiting, not Kubernetes failing. The answer is nearly always in what the process printed before it died. kubectl logs shows the current run; --previous shows the one that crashed.
What you'll do
- Find out why the pods keep restarting
- Get 2 orders pods ready and staying up
- Bonus star: Read the crashed run's output with kubectl logs --previous
The mission checks the cluster, not your keystrokes: any fix that works, passes.
Read first
More REST API missions
- 1. Config, not code
- 2. Read at start
- 3. CrashLoopBackOff
- 4. Ready when it's ready
- 5. Out of memory
- 6. Launch-day traffic
- 7. Rotate the password