Contemplating upgrading my proxmox clusterby a major revision, which
unfortunately basically involves destroying and recreating each node. A
major PITA.

The gluster bricks are all sitting on top of ZFS, so its easy enough to
preserve their data across the reinstall, but I will have to reinstall
gluster and recreate the setup (peers, datastore etc).

Can I safely recreate the bricks on top of the existing datastructure and
avoid restoring/resyncing all the data (2.5TB). I'm worried about
inadvertently creating a splitbrain problem.

If its risky I can just create a empty gluster datastore, then move the
data onto one node and allow it to sync to the others over the weekend,
only 1GB Ethernet unfortunately :(

-- 
Lindsay
_______________________________________________
Gluster-users mailing list
[email protected]
http://www.gluster.org/mailman/listinfo/gluster-users

Reply via email to