* [linux-lvm] pvmove hung at 100%
@ 2008-03-30 0:14 Chris Black
2008-04-01 16:57 ` chris
0 siblings, 1 reply; 3+ messages in thread
From: Chris Black @ 2008-03-30 0:14 UTC (permalink / raw)
To: linux-lvm
I did a pvmove between two PVs of equal size (both 4TB on different
raids) and it seems to have finished (shows 100%) yet pvdisplay commands
show that the source is still allocated and lvdisplay -m on the LV that
used the source PV shows that logical extents are still on the logical
volume pvmove0. My goal is to get this PV and the other PV this LV uses
(lv03) totally off the system and disconnected.
Does anyone have any advice/clues on how to resolve/unstick this?
Details:
Source disk is /dev/mapper/VOL04_4TB which is /dev/dm-1
Destination disk is /dev/mapper/FN00_01_4TB which is /dev/dm-10
Both are exactly 953673 PEs, source was totally allocated before pvmove
and destination was totally empty.
The command I used was:
pvmove -v -i 900 /dev/mapper/VOL04_4TB /dev/mapper/FN00_01_4TB
Subsequent pvmoves still show /dev/dm-1 100%.
lvs -av:
Finding all logical volumes
LV VG #Seg Attr LSize Maj Min KMaj KMin Origin Snap%
Move Copy% Log LV UUID
bioinfo vg01 4 -wi-ao 10.56T -1 -1 253
16
1QbcmS-tt8J-zuMB-DCUn-dAAI-pVFi-HPBUWA
lv03 vg01 2 -wI-ao 4.55T -1 -1 253
14
OXgokc-vqd7-Bcgm-8ccF-vgUo-Awfu-CNmFMo
pipeline vg01 1 -wi-ao 500.00G -1 -1 253
17
dQ3JCe-vU2m-VS7C-2VTy-nXZ4-aGdk-kbpYd6
[pvmove0] vg01 1 p-C-ao 3.64T -1 -1 253 15
/dev/dm-1 100.00 dvl7Gn-sgWk-m3Wl-hGKM-Ow6E-7j51-394T6r
rnd vg01 2 -wi-ao 7.28T -1 -1 253
18
HWBhgP-1plT-d1fu-pGdH-7Ii6-gEQN-hpu4k2
system vg01 1 -wi-ao 220.00G -1 -1 253
13
0LRNrB-TrZV-iK64-dm5C-pMTb-e8Dh-8wAVLy
lvdisplay -m /dev/vg01/lv01 shows segments:
--- Segments ---
Logical extent 0 to 953672:
Type linear
Logical volume pvmove0
Logical extents 0 to 953672
Logical extent 953673 to 1192090:
Type linear
Physical volume /dev/dm-2
Physical extents 0 to 238417
pvdisplay of /dev/dm-1 and /dev/dm-10 (source and destination) show both
as full.
Kernel version: 2.6.18.8-0.5 (SuSE 10.2 on x64)
LVM version: 2.02.13 (2006-10-27)
Library version: 1.02.12 (2006-10-13)
Driver version: 4.7.0
Thanks in advance,
Chris
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [linux-lvm] pvmove hung at 100%
2008-03-30 0:14 [linux-lvm] pvmove hung at 100% Chris Black
@ 2008-04-01 16:57 ` chris
2008-04-02 20:38 ` chris
0 siblings, 1 reply; 3+ messages in thread
From: chris @ 2008-04-01 16:57 UTC (permalink / raw)
To: linux-lvm
> I did a pvmove between two PVs of equal size (both 4TB on different
> raids) and it seems to have finished (shows 100%) yet pvdisplay commands
> show that the source is still allocated and lvdisplay -m on the LV that
> used the source PV shows that logical extents are still on the logical
> volume pvmove0. My goal is to get this PV and the other PV this LV uses
> (lv03) totally off the system and disconnected.
> Does anyone have any advice/clues on how to resolve/unstick this?
This is still stuck after two days. I am thinking perhaps I should do a
pvmove --abort to reverse the move and then do the pvmove differently in
terms of specifying the LV and multiple PVs to use instead. Does this
sound like a relatively safe course of action? I do have backups, but
would prefer not to go through all that. Also, even though I am at this
odd "pvmove 100.0% but still showing pvmove0 LV" status the data is
accessible.
Thanks in advance,
Chris
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [linux-lvm] pvmove hung at 100%
2008-04-01 16:57 ` chris
@ 2008-04-02 20:38 ` chris
0 siblings, 0 replies; 3+ messages in thread
From: chris @ 2008-04-02 20:38 UTC (permalink / raw)
To: chris; +Cc: linux-lvm
>> I did a pvmove between two PVs of equal size (both 4TB on different
>> raids) and it seems to have finished (shows 100%) yet pvdisplay
>> commands
>> show that the source is still allocated and lvdisplay -m on the LV
>> that used the source PV shows that logical extents are still on the
[stuff deleted]
>
> This is still stuck after two days. I am thinking perhaps I should do a
> pvmove --abort to reverse the move and then do the pvmove differently in
> terms of specifying the LV and multiple PVs to use instead. Does this
> sound like a relatively safe course of action? I do have backups, but
[stuff deleted]
In case anyone is curious or finds this via search later: the pvmove
--abort was successful and returned everything to the state it was before
I started the pvmove. I will be doing a different pvmove with the same
goal soon.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2008-04-02 20:38 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-03-30 0:14 [linux-lvm] pvmove hung at 100% Chris Black
2008-04-01 16:57 ` chris
2008-04-02 20:38 ` chris
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.