|
I need to switch out some old Ceph cluster hardware with new (in a somewhat unusual scenario), upgrade my Hammer cluster to Jewel, and do the chown that’s part of the upgrade. I have a notion of the process, but would like a sanity check
if people don’t mind. I’m going to be spinning up some new storage nodes for my cluster, which will ultimately replace some fairly old hardware. Because the disks in the old nodes are considerably newer than the rest of the old hardware, I’ll be evacuating
their data, removing them from the cluster, moving them to the new nodes, then re-adding them. I don’t know if this matters at all, but I’m trying to provide a complete picture. My current plan is to upgrade everything to Jewel using the “setuser match path = /var/lib/ceph/$type/$cluster-$id” ceph.conf option that allows the daemons to continue to run as root. Next, I’ll spin up the new nodes and do the evacuate/remove/physical move/re-add/repopulate dance. What fun. This should set the ownership of the contents of those OSDs to ceph:ceph. Finally, on the remaining node, one at a time I’ll bring down each daemon, chown the daemon’s data to ceph:ceph, then bring it back up. It’s a relatively small cluster, so I’d rather not bring down an entire node for the chown if I can
avoid it. Once all of that is done, I’ll remove the setuser config option. Is this a reasonable approach? Any major snags you can see, or suggestions for improving the process? Thanks. ----- Edward Huyer School of Interactive Games and Media Golisano 70-2373 152 Lomb Memorial Drive Rochester, NY 14623 585-475-6651 Obligatory Legalese: The information transmitted, including attachments, is intended only for the person(s) or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of,
or taking of any action in reliance upon this information by persons or entities other than the intended recipient is prohibited. If you received this in error, please contact the sender and destroy any copies of this information. |
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
