Bug#396582: Some additional info
Dan Pascu
dan at ag-projects.com
Fri Nov 3 22:38:18 CET 2006
martin f krafft wrote:
> also sprach Dan Pascu <dan at ag-projects.com> [2006.11.02.0946 +0100]:
>
>> Yes. In my case, if I fail a drive, it is still there in a failed state,
>> but if I then stop the raid array, when it's restarted, the failed drive
>> is no longer there, as if it was removed meanwhile, only that I never
>> issued the remove command. And when the array starts, it shows that it
>> started degraded with only 1 out of 2 drives.
>>
>
> I managed to reproduce it; you just have to write to the array after
> fail and before stop:
>
> piper:~# mdadm -Cl1 -n2 /dev/md99 /dev/sd[ef]1
> piper:~# mdadm --fail /dev/md99 /dev/sde1
> mdadm: set /dev/sde1 faulty in /dev/md99
> piper:~# dd if=/dev/zero of=/dev/md99
> [...]
> 65667072 bytes (66 MB) copied, 2.57956 seconds, 25.5 MB/s
> piper:~# mdadm -Ss
> mdadm: stopped /dev/md99
> piper:~# mdadm -As
> mdadm: /dev/md/99 has been started with 1 drive (out of 2).
> mdadm: /dev/md/99 already active, cannot restart it!
> mdadm: /dev/md/99 already active, cannot restart it!
> [...]
>
>
Martin,
I'm pretty sure that writing to the failed array is not the cause of the
problem, but more like a trigger. In my case I have never wrote to the
failed arrays during my testing. There was an array I have never wrote
anything to, and in another case there was an array I only wrote to
while it was complete (before failing it).
But I'm glad you were able to at least see the problem I'm experiencing.
One thing that intrigues me is why in my case when failing a drive and
stopping the array, after restarting it, the failed drive was already
removed (even though I never removed it myself) and the arrays started
degraded with 1 drive out of 2, and in your case the array started with
the failed drive included and reported that it started with 2 drives.
--
Dan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.alioth.debian.org/pipermail/pkg-mdadm-devel/attachments/20061103/b53096b7/attachment.htm
More information about the pkg-mdadm-devel
mailing list