Find your next open-source project

Search by name, purpose, or keyword — or paste a GitHub repository URL
All skills

gke-workload-troubleshooting

Diagnose GKE pod failures like CrashLoopBackOff, OOMKilled, and ImagePullBackOff, then propose manifest fixes.

  • Part of google/skills
  • 5.9k installs
  • DevOps & Self-hosting · Containers and clusters

When to use it

Your GKE pods keep restarting or won't start and you have logs and events but no clear root cause yet.

For example

You say

Use gke-workload-troubleshooting on my checkout-api deployment in the default namespace of cluster prod-east; pods are in CrashLoopBackOff with exit code 137.

You get
  • Root cause analysis
  • Proposed manifest patch
  • Diagnostic command sequence

How it works

  • Classifies OOM kills as container-level, node-level, or invisible cgroup v1 kills before suggesting a fix.
  • Stays read-only: it proposes gcloud or manifest changes for you to run rather than applying them.
  • If the cluster is unreachable, it hands you the exact kubectl commands and a fix based on reported symptoms.