Question
Gitlab CI/CD deployment upgrade in managed Kube fails with ImagePullBackOff
- Posted December 16, 2019
- DigitalOcean Managed Kubernetes
Hi. I’m running managed Kube with the latest version. When I do a CI/CD deployment from Gitlab, everything is green on Gitlab’s side, the deployment fires, but it will never successfully roll out in Kube. It stays stuck with an ImagePullBackOff error, unauthorized: HTTP Basic: Access denied. On the Kube side, one pod is upgraded, a fourth pod is created that displays this error, and the other two pods remain untouched.
If I manually delete the deployment before running the Gitlab pipeline, everything works. I’m at a loss for where I’m going wrong. I upgraded to more memory just to ensure it wasn’t a memory issue (I’m running basic nodes), but the problem persisted.
Subscribe
Share
Submit an answer
You can type!ref in this text area to quickly search our full set of tutorials, documentation & marketplace offerings and insert the link!
Sign in to Answer
I moved over to Helm for deployments, but the problems persist. In both my Go application and my NGINX static content server, if I do not manually run
kubectl delete deployment, the upgrade fails. I am forced to deal with application downtime. The entire point of rollingUpdate is to avoid that and I am at a loss for why it is failing.