[OpenAFS] corrupt volume -- ideas?

[Derrick J Brashear]

On Wed, 21 May 2003, Tim Czerwonka wrote:

> SalvageLog:
> @(#)Base configuration afs3.6 2.39
> 05/21/2003 11:43:00 STARTING AFS SALVAGER 2.4 (/usr/afs/bin/salvager /vicepa 2004087741 -orphans attach)
> 05/21/2003 11:43:10 Scanning inodes on device /dev/rdsk/c0t1d0s0...
> 05/21/2003 11:43:27 CHECKING CLONED VOLUME 2004087743.
> 05/21/2003 11:43:27 p.test.nobak3g.backup (2004087743) updated 07/07/2002 11:32
> 05/21/2003 11:43:27 SALVAGING VOLUME 2004087741.
> 05/21/2003 11:43:27 p.test.nobak3g (2004087741) updated 07/07/2002 11:32
> 05/21/2003 11:43:27 Cannot attach orphaned files and directories: Root directory not found
> 05/21/2003 11:43:27 Found 11 orphaned files and directories (approx. 3418510 KB)
> 05/21/2003 11:43:27 Salvaged p.test.nobak3g (2004087741): 11 files, 3418510 blocks
>
>
> Anyone have ideas for addressing the "Root directory not found" message?
>
> The volume resides on a sol 8 transarc AFS fileserver, but the same
> results are found when the volume has been moved to an OpenAFS fileserver.

The volume's root vnode got trashed somehow, and the salvager doesn't know
how to recover.