9:25 a.m.
On Tue, Jan 30, 2018 at 3:14 PM, Elad Ben Aharon <ebenahar(a)redhat.com>
wrote:
In a case of disk migration failure with leftover LV on the
destination
domain, lvremove is what needed. Also, make sure to remove the image
directory on the destination domain (located under
/rhev/data-center/%spuuid%/%sduuid%/images/)
Ok, exactly the 2 steps I've done after noticing the broken link inside the
image directory. I suspected something to be done at rdbms level.
Thanks for confirmation.
The bad was that the VM crashed because of this error.
Not a big problem in my case, as this is a test env.
I had to move many disk images from SAN to SAN, about 200Gb size each, and
so I powered off the VMs and then moved the disks.
I didn't get anymore the problem at destination, but I did get sometimes
the opposite: unable to remove the LV at source:
command HSMGetAllTasksStatusesVDS failed: Cannot remove Logical Volume:
(['Cannot remove Logical Volume:
(u\'f7b481c8-b744-43d4-9faa-0e494a308490\',
"[\'6d23ec3e-79fb-47f0-85ff-08480476ea68\']")'],)
and I see that the apparently the LV remains active/open (ao flag) at
source...
Even after putting the source storage domain, by now empty, into
maintenance, at host side I get this kid of thing:
# lvs f7b481c8-b744-43d4-9faa-0e494a308490
LV VG
Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert
45e20ac8-f3c6-462b-a2c3-30f0c8262f84 f7b481c8-b744-43d4-9faa-0e494a308490
-wi-ao---- 200.00g
4d27c0e1-546b-48ba-80fb-95dcd87b5583 f7b481c8-b744-43d4-9faa-0e494a308490
-wi-ao---- 380.00g
675d5f06-320a-4236-8d57-9ff7cc7eb200 f7b481c8-b744-43d4-9faa-0e494a308490
-wi------- 50.00g
6d23ec3e-79fb-47f0-85ff-08480476ea68 f7b481c8-b744-43d4-9faa-0e494a308490
-wi-ao---- 200.00g
823a79f2-b09a-4665-9dfa-8ccd2850225f f7b481c8-b744-43d4-9faa-0e494a308490
-wi------- 128.00m
863062c3-b3b0-4494-ad2f-76a1c29c069a f7b481c8-b744-43d4-9faa-0e494a308490
-wi------- 128.00m
a2d5f043-55c2-4701-ac8c-ccb796a799d1 f7b481c8-b744-43d4-9faa-0e494a308490
-wi-ao---- 200.00g
e26eac74-4445-4f3b-b2f5-ef048ae6d91b f7b481c8-b744-43d4-9faa-0e494a308490
-wi-ao---- 200.00g
ids f7b481c8-b744-43d4-9faa-0e494a308490
-wi-a----- 128.00m
inbox f7b481c8-b744-43d4-9faa-0e494a308490
-wi-a----- 128.00m
leases f7b481c8-b744-43d4-9faa-0e494a308490
-wi-a----- 2.00g
master f7b481c8-b744-43d4-9faa-0e494a308490
-wi------- 1.00g
metadata f7b481c8-b744-43d4-9faa-0e494a308490
-wi-a----- 512.00m
outbox f7b481c8-b744-43d4-9faa-0e494a308490
-wi-a----- 128.00m
xleases f7b481c8-b744-43d4-9faa-0e494a308490
-wi-a----- 1.00g
and for some hours (now I don't get it anymore, perhaps a consequence of
putting the SD into maintenance), the lvs command exhibited also this kind
of output/error
WARNING: PV lp71QQ-9ozC-l2Yx-o9mS-ikpg-c8ft-DfUTyK on
/dev/c0097b1a-a387-4ffa-a62b-f9e6972197ef/e26eac74-4445-4f3b-b2f5-ef048ae6d91b
was already found on
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/e26eac74-4445-4f3b-b2f5-ef048ae6d91b.
WARNING: PV nRe20Y-XuD7-aFSl-INgl-WlkO-JGhK-zekrOl on
/dev/c0097b1a-a387-4ffa-a62b-f9e6972197ef/45e20ac8-f3c6-462b-a2c3-30f0c8262f84
was already found on
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/45e20ac8-f3c6-462b-a2c3-30f0c8262f84.
WARNING: PV NjcwGm-Rf1H-NV9p-eCjU-qmF8-WCkB-CdzLmN on
/dev/c0097b1a-a387-4ffa-a62b-f9e6972197ef/6d23ec3e-79fb-47f0-85ff-08480476ea68
was already found on
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/6d23ec3e-79fb-47f0-85ff-08480476ea68.
WARNING: PV 0dmwuU-1v6k-8weJ-yEg3-4Oup-w4Bp-GlNeRf on
/dev/c0097b1a-a387-4ffa-a62b-f9e6972197ef/a2d5f043-55c2-4701-ac8c-ccb796a799d1
was already found on
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/a2d5f043-55c2-4701-ac8c-ccb796a799d1.
WARNING: PV lp71QQ-9ozC-l2Yx-o9mS-ikpg-c8ft-DfUTyK prefers device
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/e26eac74-4445-4f3b-b2f5-ef048ae6d91b
because device is used by LV.
WARNING: PV nRe20Y-XuD7-aFSl-INgl-WlkO-JGhK-zekrOl prefers device
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/45e20ac8-f3c6-462b-a2c3-30f0c8262f84
because device is used by LV.
WARNING: PV NjcwGm-Rf1H-NV9p-eCjU-qmF8-WCkB-CdzLmN prefers device
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/6d23ec3e-79fb-47f0-85ff-08480476ea68
because device is used by LV.
WARNING: PV 0dmwuU-1v6k-8weJ-yEg3-4Oup-w4Bp-GlNeRf prefers device
/dev/f7b481c8-b744-43d4-9faa-0e494a308490/a2d5f043-55c2-4701-ac8c-ccb796a799d1
because device is used by LV.
I have been able to power on VMs with the new underlying storage domain
active.
Any hint on how to clean. I can shutdown and restart this host (that in
this moment is the only one running), but it could be of help in other
similar scenarios too...
Gianluca