Pod CrashLoopBackOff on new deployment

[geek]

Deployed new version of app, but pods enter CrashLoopBackOff with "connection refused". Logs show DB timeout. Ideas?

[Andrea]

Likely your init container finishes before the DB is actually ready. Try adding startupProbe or delay.

[geek]

Yeah, increasing readiness delay fixed it. Thanks!

[Paul Davis]

Trying to build Docker images inside GitLab CI runners using Docker-in-Docker. Getting network issues and rate limiting from Docker Hub.

[geek]

Use `--network host` and set up DockerHub mirror cache like JFrog or Harbor. Also enable auth token for rate limit bypass.

[Paul Davis]

Setting up Harbor helped. Thanks!