[OpenAFS] Re: Tracking down AFS Fileserver corruption

Andrew Deason adeason@sinenomine.net
Tue, 29 Nov 2011 10:50:27 -0600

On Mon, 28 Nov 2011 13:13:32 -0500
Jack Neely <jjneely@pams.ncsu.edu> wrote:

> and the volume being moved at that exact time would be corrupt.  Usually
> the server would soon detect this and salvage the volume, but the level
> of corruptions has varied.

"Corruption". What, exactly? What do the salvage logs say after the
fact, and what did the fileserver or volserver logs say was wrong about
the volume that required salvaging?

> The above messages and corruption only seem to happen when volume moves
> are in progress.  Running fdisk -l on an idle server produces no
> messages.

The fileserver and volserver just do normal userspace i/o. If you can
just perform some i/o on a similarly-configured device and filesystem
and get a similar thing to happen, that's probably an easier test case
for RedHat or whomever to look at. Maybe just fiddle around with 'dd' or
run a filesystem benchmark or something.

Andrew Deason