[pve-devel] Fatal bug in migration

[Michael Rasmussen]

Hi all,

Previous enterprise and the version released yesterday have a fatal bug
which affects VM's using storage zfs_over_iscsi. Bug means that every
VM online migrated will crash on the new node!!! The bug was introduced
with the new way of doing migration.

Every migration shows this:
Jan 03 21:26:17 starting migration of VM 100 to node 'esx1' (10.0.0.1)
Jan 03 21:26:17 copying disk images
Jan 03 21:26:17 starting VM 100 on remote node 'esx1'
Jan 03 21:26:23 starting online/live migration on tcp:10.0.0.1:60000
Jan 03 21:26:23 migrate_set_speed: 8589934592
Jan 03 21:26:23 migrate_set_downtime: 0.1
Jan 03 21:26:23 set migration_caps
Jan 03 21:26:23 set cachesize: 107374182
Jan 03 21:26:23 spice client_migrate_info
Jan 03 21:26:23 start migrate command to tcp:10.0.0.1:60000
Jan 03 21:26:25 migration speed: 512.00 MB/s - downtime 71 ms
Jan 03 21:26:25 migration status: completed
Jan 03 21:26:26 ERROR: VM 100 not running
Jan 03 21:26:26 ERROR: command '/usr/bin/ssh -o 'BatchMode=yes' root at 10.0.0.1 qm resume 100 --skiplock --nocheck' failed: exit code 2
Jan 03 21:26:26 Waiting for spice server migration
Jan 03 21:26:27 ERROR: migration finished with problems (duration 00:00:10)
TASK ERROR: migration problems

-- 
Hilsen/Regards
Michael Rasmussen

Get my public GnuPG keys:
michael <at> rasmussen <dot> cc
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E
mir <at> datanom <dot> net
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C
mir <at> miras <dot> org
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917
--------------------------------------------------------------
/usr/games/fortune -es says:
  Cum tacent, clamant. When they are silent, they shout. -Cicero
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 195 bytes
Desc: OpenPGP digital signature
URL: <http://lists.proxmox.com/pipermail/pve-devel/attachments/20170103/52ccc61d/attachment.sig>