thank you, xfs team

thank you, xfs team

[Hungerburg]

hello there,

I hope this is not the wrong place, yet I have to write this now, before
I forget about it:

I had to swap a dying harddrive. after receiving many mails from the
smart daemon, recent kernel patches to xfs made me aware of the full
extent of the problem.

keywords:
- XFS internal error XFS_WANT_CORRUPTED_RETURN
- Device: /dev/hda, 1 Currently unreadable (pending) sectors
- Device: /dev/hda, 1 Offline uncorrectable sectors

the tools to copy what is still there (data) are a boon - they require
some reading, but the thing is doable (xfs_copy to good disk, then
xfs_dump from there)

yours

Hungerburg

-- 
||| | | | | |  |  |  |   http://arton.cunst.net/   |  |  |  | | | | | |||

Re: thank you, xfs team

[Nathan Scott]

On Tue, Jul 04, 2006 at 12:11:17AM +0200, Hungerburg wrote:
> hello there,
> 
> I hope this is not the wrong place, yet I have to write this now, before
> I forget about it:

xfs@oss.sgi.com is the best place.

> I had to swap a dying harddrive. after receiving many mails from the
> smart daemon, recent kernel patches to xfs made me aware of the full
> extent of the problem.

Hmm, the second part of that sentence doesn't make sense to me.

> keywords:
> - XFS internal error XFS_WANT_CORRUPTED_RETURN
> - Device: /dev/hda, 1 Currently unreadable (pending) sectors
> - Device: /dev/hda, 1 Offline uncorrectable sectors

This looks like expected behaviour - XFS shut down the filesystem on
detecting IO errors on filesystem metadata blocks.

> the tools to copy what is still there (data) are a boon - they require
> some reading, but the thing is doable (xfs_copy to good disk, then
> xfs_dump from there)

OK, good to hear.

cheers.

-- 
Nathan