Bug#819171: linux-image-3.16.0-4-amd64: lvm cache cachepool kernel regression, lvm stops working when upgrading from 3.16.7-ckt11-1 to 3.16.7-ckt20-1
Hi Jelle
What did you do to this poor system?
On Thu, Mar 24, 2016 at 02:08:10PM +0100, Jelle de Jong wrote:
> [ 27.687179] md/raid1:md0: active with 1 out of 2 mirrors
> [ 27.693133] md0: detected capacity change from 0 to 1000194564096
> mdadm: /dev/md0 has been started with 1 drive (out of 2).
One device missing.
> [ 27.728351] md/raid1:md1: not clean -- starting background reconstruction
> [ 27.735933] md/raid1:md1: active with 2 out of 2 mirrors
> [ 27.741938] md1: detected capacity change from 0 to 1000194564096
> mdadm: /dev/md1 has been started with 2 drives.
Unclean.
> [ 29.599320] md: resuming resync of md6 from checkpoint.
> [ 29.652105] md: resuming resync of md1 from checkpoint.
Resume from checkpoint, so a rebuild was already running.
> [ 45.642884] md/raid1:mdX: active with 2 out of 2 mirrors
> [ 45.648979] mdX: invalid bitmap file superblock: bad magic
> [ 45.655123] mdX: failed to create bitmap (-22)
> [ 45.660202] device-mapper: table: 251:7: raid: Fail to run raid array
> [ 45.667396] device-mapper: ioctl: error adding target to table
And another raid where just the bitmap is toast.
You could try removing the bitmap with mdadm --grow --bitmap=none.
Bastian
--
Warp 7 -- It's a law we can live with.
Reply to: