[OpenAFS] Volumes that need to be salvaged

[Esther Filderman]

On 9/5/05, Jason Mader <jason+openafs@ncac.gwu.edu> wrote:
> What is the procedure to fix the following?
>=20
> @(#) OpenAFS 1.4.0-rc3 built  2005-09-05
> 09/05/2005 10:49:23 STARTING AFS SALVAGER 2.4 (/usr/local/libexec/
> openafs/salvager)
> 09/05/2005 10:49:23 Starting salvage of file system partition /vicepa
> 09/05/2005 10:49:23 SALVAGING FILE SYSTEM PARTITION /vicepa
> (device=3Ddks0d2s7)
> 09/05/2005 10:49:29 Wrong size (2048 instead of 2560) in inode file
> for <null>
> 09/05/2005 10:49:29 *** I/O error 1009 when writing a tmp inode file /
> vicepa/salvage.inodes.dks0d2s7.303; Not salvaged dks0d2s7 ***
> Increase space on partition or use '-tmpdir'
>=20
> Mon Sep  5 10:49:30 2005 XFS/EFS File server starting
> Mon Sep  5 10:49:30 2005 Set thread id 14 for FSYNC_sync
> Mon Sep  5 10:49:30 2005 Partition /vicepa: attaching volumes
> Mon Sep  5 10:49:30 2005 VAttachVolume: volume /vicepa/
> V0536870927.vol needs to be salvaged; not attached.

/vicepa is either full or has a [physical] bad spot on it.  Try moving
some other volumes from it and then salvaging the volume ["bos salvage
yourservername a 536870927" will do it without taking the rest of the
fileserver down].

Check your system log{,s} for reports of disk problems.