Details
-
Bug
-
Resolution: Done
-
Critical
-
None
-
6.5 SP1, 7.0, 7.1
-
None
Description
After applying last updates (XS70E015 and upper), then rebooting one host after another (live migrate between host), I found *all* my VMs got their disk missing (no VDI, no VBD attached). Running VMs were still able to write, so the share was still working. Reboot a VM fails because no disk attached anymore.
They were all on one SR. When I check this SR, I got "0 VDIs" (no disk visible in xe no Xen Orchestra nor XenCenter. But it's correctly mounted in all hosts, and ls -la inside the mount list show all VHD files and a filelog.txt
I tried to rescan, disconnect/reconnect the SR, forget and re-attached it, restart the XAPI, reboot hosts: same result. 0x VDIs.
I don't know what do to (this is a lab so not critical but I know a lot of people using NFS shares).
- Is there a way to check what's wrong with this SR? Something I should find in the file list that I didn't have?
- Can I force XenServer to read those VHD files again?
- Could it be a XAPI DB corruption because of a disk problem on the master that break the metadata? (link between VHD files and the disk UUIDs etc.)