contact problem? (was Re: Spurious Filesystem corruption with ext3 + large (<400GB) hw RAID ...)

contact problem? (was Re: Spurious Filesystem corruption with ext3 + large (<400GB) hw RAID ...)

[Manfred Wassmann]

On Fri, Feb 6, 2009 at 7:50 PM, Ray Lee <ray-lk@madrabbit.org> wrote:

> Huh. It may be related to whatever kernel version Ubuntu uses to boot
> up the install media. Try a different version of Ubuntu (older, newer
> if a newer one exists) to see if it has the same problem.

Thank you for your notice but as I mentioned the problem occurred
first with my custom built kernel and the Ubuntu installation disk was
used only to reproduce the error -- Ubuntu is not what I use anyhow,
I'm on Debian since I switched from Slackware in 1995 ;-)

But now it looks like we had an exotic hardware problem here. Just
before the weekend the RAID controller reported a degraded array and
switched to the hotspare disk. On Monday the original array was
recreated and everything worked fine until Tuesday morning when the
array again was degraded. But then since the disk was checked and
reinserted into the array all problems are gone :-\

Is it possible that such a problem is caused by bad contact within the
drive bay connector?

The reason why it occurred with ext3 only might then be that ext3
stores it's backup superblocks at addresses which are otherwise unused
by a largely empty filesystem.

regards Manfred
-- 
Unix "Birthday" on 2009-02-13 23:31:30 UTC
...it's 1234567890 seconds since the epoch.

Re: contact problem? (was Re: Spurious Filesystem corruption with ext3 + large (<400GB) hw RAID ...)

[Ray Lee]

On Fri, Feb 13, 2009 at 8:29 AM, Manfred Wassmann
<tux.wassmann@googlemail.com> wrote:
> On Fri, Feb 6, 2009 at 7:50 PM, Ray Lee <ray-lk@madrabbit.org> wrote:
>
>> Huh. It may be related to whatever kernel version Ubuntu uses to boot
>> up the install media. Try a different version of Ubuntu (older, newer
>> if a newer one exists) to see if it has the same problem.
>
> Thank you for your notice but as I mentioned the problem occurred
> first with my custom built kernel and the Ubuntu installation disk was
> used only to reproduce the error -- Ubuntu is not what I use anyhow,
> I'm on Debian since I switched from Slackware in 1995 ;-)
>
> But now it looks like we had an exotic hardware problem here. Just
> before the weekend the RAID controller reported a degraded array and
> switched to the hotspare disk. On Monday the original array was
> recreated and everything worked fine until Tuesday morning when the
> array again was degraded. But then since the disk was checked and
> reinserted into the array all problems are gone :-\
>
> Is it possible that such a problem is caused by bad contact within the
> drive bay connector?
>
> The reason why it occurred with ext3 only might then be that ext3
> stores it's backup superblocks at addresses which are otherwise unused
> by a largely empty filesystem.

(Ah, I missed this message first time around. Please always do a
reply-to-all for lkml.)

Bad contacts, sure, or (more likely) a marginal power supply? Can you
try a different enclosure, or a different power supply for the
enclosure?

Re: contact problem? (was Re: Spurious Filesystem corruption with ext3 + large (<400GB) hw RAID ...)

[Manfred Wassmann]

On Sat, Feb 14, 2009 at 18:40, Ray Lee <madrabbit@gmail.com> wrote:
[...]
> (Ah, I missed this message first time around. Please always do a
> reply-to-all for lkml.)

Thanks, that's always a problem with mailing lists, you never know,
some people want CCs and some -- like me -- don't ;-)

> Bad contacts, sure, or (more likely) a marginal power supply? Can you
> try a different enclosure, or a different power supply for the
> enclosure?

Now it looks like the disk is the problem, this week the array again
reported degrades state with the same disk as before being flagged as
defective but some time later everything was OK. The disk now has been
replaced.
Live was easier when disks were either good or bad ;-)