Search results

Edit: FYI: I successfully replaced a different disk on a different node earlier today, and that's still rebalancing. I had an issue destroying the old drive, but thought that might be normal on a failed disk: destroy OSD osd.0 Remove osd.0 from the CRUSH map Remove the osd.0 authentication...

Mira, Thanks for the reply, I just finished replacing a node a on a test cluster and it looks like replacing a existing node would work better for me in production. I ran into issues with the monitor service for Ceph not wanting to start correctly. It appears that the correct thing to do...

Greetings All, We're running a five node cluster and two of the members are getting old and we want to replace them. Is the "best practice" to, e.g., delete node 2 from the cluster and make a new node 2 and join that to the cluster, or should we add a node #6 and then delete #2? This...