Critical Troubleshooting Tools for Rook Clusters
summary
A kubectl krew plugin called rook-ceph installs via krew and exposes one-line commands: rook status (prints Ceph and CR status and capacity), purge-osd (removes a failed OSD without creating a separate job), and a debug deployment flow that scales down the original pod, creates a debug pod with operator-avoid labels and optional alternate image, lets operators run mon/osd tools, then restores the original. Health check enforces three mons/OSDs on different nodes; restore quorum is a single confirmable command. Demo runs on Minikube; scripts move from shell to Go.
AI-written summary · about
chapters (11)
- 0:00Introduction to troubleshooting
- 1:27What is Krew and Rook Safe
- 3:12Installing Krew and Rook Safe
- 9:31Demo overview
- 12:21Basic Rook Safe commands
- 14:35Debug mode deep dive
- 17:50Purge OSD command
- 20:38Health check command
- 22:05Mon quorum restore
- 26:06Future work and references
- 27:29Q&A session
