In experimenting with different storage and disk types with ceph I have 88 OSDs across 3 nodes. When I attempt to bulk stop I just get a small UI Bulk Stop screen with a red border and nothing within it to select. I've also found that multi select is not an option either from the OSD list in the webUI. Is there a bug with the interface (5.3-6) or an alternate command set from console to bulk stop OSDs? Otherwise that's a carpel tunnel inducing process to stop them and then remove/destroy them.
[SOLVED] Ceph Bulk Actions
- Thread starter bensode
- Start date
-
- Tags
- ceph osd bulk stop
The "Bulk Stop" menu in the GUI is for Guest-Systems (it does not vanish if you leave the Ceph menu) not for OSDs.
Currently we don't have a "Bulk Stop" implemented in the GUI for OSDs (out of curiosity, when would you want to restart all OSDs, but not the host?).
You can stop ceph-osd.target via systemctl on the command line- that should take all OSDs down.
Currently we don't have a "Bulk Stop" implemented in the GUI for OSDs (out of curiosity, when would you want to restart all OSDs, but not the host?).
You can stop ceph-osd.target via systemctl on the command line- that should take all OSDs down.
Thanks I'll give that command line a shot. I'm new to PVE and also Ceph so I'm experimenting with different physical configurations of disks to better understand them. There's a lot of material to cover with both products and the Ceph portion is definitely information overload.
Can relate to that
- and the learning doesn't stop for a long time - as a payoff you can get a really solid distributed storage!Good luck and happy hacking!