* RAID 0, disks swapped ?
@ 2008-10-17 11:18 c.baegert-listes
2008-10-17 12:36 ` Mario 'BitKoenig' Holbe
2008-10-17 13:18 ` Bill Davidsen
0 siblings, 2 replies; 10+ messages in thread
From: c.baegert-listes @ 2008-10-17 11:18 UTC (permalink / raw)
To: linux-raid
Hi,
I had a RAID 10, crashed since a few days, made with a soft RAID 0 (/dev/md5)
of 2 soft RAID 1 (/dev/md3 and /dev/md4). I'm using default kernel on Debian
Etch.
I had to recreate the main RAID 0 (mdadm --create), I thought that the order
of /dev/md3 and /dev/md4 was not important (mdadm is known to find the good
order automatically), but it seems it took the 2 arrays in the wrong order :
when I tried to mount it, it failed, with the message "you need to
reiserfsck", so I made a fsck, it said "you have to rebuild-tree", I did it
and now all my files are messed up, and it seems theye are sliced !!!
I asked for a Reiserfs specialist who said my whole partition has an offset of
64K. And it's the chunk size of my RAID 0 !
So do you think that the order of my sub-arrays REALLY changed in the RAID 0 ?
and do you think possible to reorder that, before retrying higher level tools
like fsck ?
Thank you for your help !
Regards,
--
Christophe
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 11:18 RAID 0, disks swapped ? c.baegert-listes
@ 2008-10-17 12:36 ` Mario 'BitKoenig' Holbe
2008-10-17 13:57 ` c.baegert-listes
2008-10-17 13:18 ` Bill Davidsen
1 sibling, 1 reply; 10+ messages in thread
From: Mario 'BitKoenig' Holbe @ 2008-10-17 12:36 UTC (permalink / raw)
To: linux-raid
c.baegert-listes@lixium.fr <c.baegert-listes@lixium.fr> wrote:
> I had to recreate the main RAID 0 (mdadm --create), I thought that the order
> of /dev/md3 and /dev/md4 was not important (mdadm is known to find the good
> order automatically),
This is true for assembling an existing RAID (provided it has a
superblock), but when you create a new one ... where should md know the
order except from what you gave it?
But... why did you have to recreate it at all?
regards
Mario
--
We are the Bore. Resistance is futile. You will be bored.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 11:18 RAID 0, disks swapped ? c.baegert-listes
2008-10-17 12:36 ` Mario 'BitKoenig' Holbe
@ 2008-10-17 13:18 ` Bill Davidsen
2008-10-17 13:48 ` c.baegert-listes
1 sibling, 1 reply; 10+ messages in thread
From: Bill Davidsen @ 2008-10-17 13:18 UTC (permalink / raw)
To: c.baegert-listes; +Cc: linux-raid
c.baegert-listes@lixium.fr wrote:
> Hi,
>
> I had a RAID 10, crashed since a few days, made with a soft RAID 0 (/dev/md5)
> of 2 soft RAID 1 (/dev/md3 and /dev/md4). I'm using default kernel on Debian
> Etch.
>
Just for educational purposes (won't help fix your problem), you had
RAID 1+0 rather than RAID-10. See the mdadm docs for the difference. And
reading old discussions here on the performance of RAID-10 with two far
copies would be useful as well.
Sorry for your loss.
--
Bill Davidsen <davidsen@tmr.com>
"Woe unto the statesman who makes war without a reason that will still
be valid when the war is over..." Otto von Bismark
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 13:18 ` Bill Davidsen
@ 2008-10-17 13:48 ` c.baegert-listes
2008-10-17 14:05 ` Mario 'BitKoenig' Holbe
0 siblings, 1 reply; 10+ messages in thread
From: c.baegert-listes @ 2008-10-17 13:48 UTC (permalink / raw)
To: Bill Davidsen; +Cc: linux-raid
Hi,
Le vendredi 17 octobre 2008, Bill Davidsen a écrit :
> Just for educational purposes (won't help fix your problem), you had
> RAID 1+0 rather than RAID-10. See the mdadm docs for the difference. And
> reading old discussions here on the performance of RAID-10 with two far
> copies would be useful as well.
Yeah, I saw them, I use directly raid10 for a few months, but this array was
older. The good point of the RAID 1+0 solution is that I just have to repair
a raid0 array now. But how ?
Regards,
--
Christophe
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 12:36 ` Mario 'BitKoenig' Holbe
@ 2008-10-17 13:57 ` c.baegert-listes
2008-10-17 14:01 ` Mario 'BitKoenig' Holbe
0 siblings, 1 reply; 10+ messages in thread
From: c.baegert-listes @ 2008-10-17 13:57 UTC (permalink / raw)
To: Mario 'BitKoenig' Holbe; +Cc: linux-raid
Le vendredi 17 octobre 2008, vous avez écrit :
> This is true for assembling an existing RAID (provided it has a
> superblock), but when you create a new one ... where should md know the
> order except from what you gave it?
> But... why did you have to recreate it at all?
If I remember well, a re-add failed after a disk lost sync, so I had to
re-create the RAID 1 array after stopping it, and then I had to stop the RAID
0 before...
Regards,
--
Christophe
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 13:57 ` c.baegert-listes
@ 2008-10-17 14:01 ` Mario 'BitKoenig' Holbe
2008-10-17 14:07 ` c.baegert-listes
0 siblings, 1 reply; 10+ messages in thread
From: Mario 'BitKoenig' Holbe @ 2008-10-17 14:01 UTC (permalink / raw)
To: linux-raid
c.baegert-listes@lixium.fr <c.baegert-listes@lixium.fr> wrote:
> re-create the RAID 1 array after stopping it, and then I had to stop the RAID
> 0 before...
Yes, stop the raid0, and start it again, but why create?
regards
Mario
--
The secret that the NSA could read the Iranian secrets was more
important than any specific Iranian secrets that the NSA could
read. -- Bruce Schneier
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 13:48 ` c.baegert-listes
@ 2008-10-17 14:05 ` Mario 'BitKoenig' Holbe
2008-10-17 14:34 ` c.baegert-listes
0 siblings, 1 reply; 10+ messages in thread
From: Mario 'BitKoenig' Holbe @ 2008-10-17 14:05 UTC (permalink / raw)
To: linux-raid
c.baegert-listes@lixium.fr <c.baegert-listes@lixium.fr> wrote:
> older. The good point of the RAID 1+0 solution is that I just have to repair
> a raid0 array now. But how ?
That's most likely gone - at least your data is most likely corrupted
after your reiserfsck and rebuild-tree, unless all you did did not write
to the disks (which is not very likely :)).
You could try to recreate the array in the correct order, perhaps you
can get a little back but i think its time to consult your backups,
alone to find out what's corrupted and what's not.
regards
Mario
--
Damit das Mögliche entsteht, muß immer wieder das Unmögliche versucht werden.
-- Hermann Hesse
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 14:01 ` Mario 'BitKoenig' Holbe
@ 2008-10-17 14:07 ` c.baegert-listes
0 siblings, 0 replies; 10+ messages in thread
From: c.baegert-listes @ 2008-10-17 14:07 UTC (permalink / raw)
To: Mario 'BitKoenig' Holbe; +Cc: linux-raid
Le vendredi 17 octobre 2008, vous avez écrit :
> c.baegert-listes@lixium.fr <c.baegert-listes@lixium.fr> wrote:
> > re-create the RAID 1 array after stopping it, and then I had to stop the
> > RAID 0 before...
>
> Yes, stop the raid0, and start it again, but why create?
Yes, I should have tried to assemble it... Sometimes man should better not go
to work and stay in his bed. But it's not what I did then.
Christophe
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 14:05 ` Mario 'BitKoenig' Holbe
@ 2008-10-17 14:34 ` c.baegert-listes
2008-10-17 17:44 ` Mario 'BitKoenig' Holbe
0 siblings, 1 reply; 10+ messages in thread
From: c.baegert-listes @ 2008-10-17 14:34 UTC (permalink / raw)
To: Mario 'BitKoenig' Holbe; +Cc: linux-raid
Le vendredi 17 octobre 2008, vous avez écrit :
> You could try to recreate the array in the correct order
Is the order the number [1] or [0] in the line below ?
md5 : active raid0 md4[1] md3[0]
Regards,
--
Christophe
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: RAID 0, disks swapped ?
2008-10-17 14:34 ` c.baegert-listes
@ 2008-10-17 17:44 ` Mario 'BitKoenig' Holbe
0 siblings, 0 replies; 10+ messages in thread
From: Mario 'BitKoenig' Holbe @ 2008-10-17 17:44 UTC (permalink / raw)
To: linux-raid
c.baegert-listes@lixium.fr <c.baegert-listes@lixium.fr> wrote:
> Is the order the number [1] or [0] in the line below ?
> md5 : active raid0 md4[1] md3[0]
Yes, this should be the result of mdadm -C ... /dev/md5 /dev/md3 /dev/md4
regards
Mario
--
There are two major products that come from Berkeley: LSD and UNIX.
We don't believe this to be a coincidence. -- Jeremy S. Anderson
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2008-10-17 17:44 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-10-17 11:18 RAID 0, disks swapped ? c.baegert-listes
2008-10-17 12:36 ` Mario 'BitKoenig' Holbe
2008-10-17 13:57 ` c.baegert-listes
2008-10-17 14:01 ` Mario 'BitKoenig' Holbe
2008-10-17 14:07 ` c.baegert-listes
2008-10-17 13:18 ` Bill Davidsen
2008-10-17 13:48 ` c.baegert-listes
2008-10-17 14:05 ` Mario 'BitKoenig' Holbe
2008-10-17 14:34 ` c.baegert-listes
2008-10-17 17:44 ` Mario 'BitKoenig' Holbe
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).