On Mon, 10 Oct 2011 22:47:58 +0100 Michael Busby <michael.a.busby@gmail.com>
wrote:

> I have a quick question i remember reading somewhere about not using
> metadata version 0.9 with drives larger than 2tb,
> > at the moment i have the following
> >
> > root@BlueBolt:~# cat /proc/mdstat
> > Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
> > md0 : active raid5 sdd[2] sde[0] sdb[3] sdc[1]
> >       5860543488 blocks level 5, 512k chunk, algorithm 2 [4/4] [UUUU]
> >       bitmap: 2/15 pages [8KB], 65536KB chunk
> > unused devices: <none>
> > root@BlueBolt:~# mdadm --detail /dev/md0
> > /dev/md0:
> >         Version : 0.90
> >   Creation Time : Mon Jul  4 15:08:38 2011
> >      Raid Level : raid5
> >      Array Size : 5860543488 (5589.05 GiB 6001.20 GB)
> >   Used Dev Size : 1953514496 (1863.02 GiB 2000.40 GB)
> >    Raid Devices : 4
> >   Total Devices : 4
> > Preferred Minor : 0
> >     Persistence : Superblock is persistent
> >   Intent Bitmap : Internal
> >     Update Time : Mon Oct 10 22:44:11 2011
> >           State : active
> >  Active Devices : 4
> > Working Devices : 4
> >  Failed Devices : 0
> >   Spare Devices : 0
> >          Layout : left-symmetric
> >      Chunk Size : 512K
> >            UUID : ddab6c38:dee3ead0:95ba4558:1c9a49ed (local to host BlueBolt)
> >          Events : 0.2836102
> >     Number   Major   Minor   RaidDevice State
> >        0       8       64        0      active sync   /dev/sde
> >        1       8       32        1      active sync   /dev/sdc
> >        2       8       48        2      active sync   /dev/sdd
> >        3       8       16        3      active sync   /dev/sdb
> > which as you can see if using 0.90, i am looking at replacing all the 2tb drives with 3tb versions, would i need to update the metadata version? if so how can i go about this?

With a really recent kernel (3.1) and recent mdadm (also not released yet),
0.90 can go up to 4TB (it has 32 bits to count kilobytes with).

Alternately you need to convert to 1.0 metadata.

Currently the only way to do this is to 'create' the array again.
Be sure to specified the same chunk size, the right metadata, the name level
and number of disks, and the correct disks in the correct order.
An use "--assume-clean".  Then check your data is still consistent.
With --assume-clean and a read-only mount, no data will actually be changed,
only metadata.

NeilBrown