[OpenAFS] Re: Move of large volume fails: Rx call failed during dump

Andrew Deason adeason@sinenomine.net
Wed, 25 Apr 2012 09:39:16 -0500

On Wed, 25 Apr 2012 16:19:30 +0200
Christopher Odenbach <odenbach@uni-paderborn.de> wrote:

> Tue Apr 24 20:57:48 2012 1 Volser: ReadVnodes: IH_CREATE: File exists
> - - restore aborted
> Tue Apr 24 21:03:04 2012 1 Volser: Delete: volume 537020184 deleted
> I have already tried to salvage the volume, but still the move fails.
> What may be the reason and how can I fix it?

Where did you salvage it? That log indicates that the volume at least
partially exists at the destination site. A salvage of volume 537020184
on the destination fileserver/partition should clear it up if there is
some cruft just lying around there. Or you can try a 'vos zap' or 'vos
zap -f' of that volume id at the destination. If a salvage at the
destination doesn't clear it up, what does the salvage log say?

If you want to check before trying that long move again, the destination
should have no files in the directory tree /vicepX/AFSIDat/M/MQY+U,
assuming you don't have any RO or BK volumes, etc for that volume. (Be
sure not to touch anything in there.)

Andrew Deason