> I have dig a little big, and I find the real problem. > > block-job-cancel is async, so it take 1 or 2s for the block-job is really > finished. > > That's why we got the "rbd has watcher" error, because the disk was always > open in qemu. > > So, we just need to check that block-job don't exist anymore before continue. > > I'll send a patch today.
Great! Btw, did you visit OpenStack Summit in Paris? If so, any interesting news? _______________________________________________ pve-devel mailing list [email protected] http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
