Feb 2026 Update:
This post is from 2021. Since then, I’ve made many changes to my cluster and Rancher has also made many changes to their product. Some better, some still challenging. I haven’t had to rebuild my cluster from scratch since then, which is a positive improvement. Now my issues are mostly due to self-hosted Kubernetes cluster + Longhorn PVC issues instead of Rancher. They deprecated RKE1 without an in-place migration plan, but I figured out how to migrate to NixOS. Every upgrade to Rancher fixes one issue, then add another. I stopped using Rancher Fleet because it was buggy and started using cdk8s+Helm. While it had it’s own issues, I was able to more easily navigate them. Rancher v2.11 broke copy from the view YAML screen. v2.13 got rid of the combined Workloads screen which pulled in deployments, jobs, etc. because of supposed performance issues. I used that feature way too much. I’ve remained on v2.11.