drain + destroy + zap + create
The 'orchestrator service-instance redeploy osd $n' works now, inasmuch as it lets you upgrade the osd code to a new container image. Is that what this ticket is for? Or are you thinking drain + destroy + zap + create (e.g., "repave")?
- Subject changed from redeploy osds to repave osds
- Description updated (diff)
Yeah. Didn't know that "repave" is something special. the Trello card was not self-explanatory. I've renamed this to "repave"
- Category changed from cephadm to cephadm/osd
Also available in: Atom