[OpenAFS] mangled volume dump

[Jeffrey Hutzelman]

On Thursday, January 20, 2005 12:12:42 -0500 "John S. Bucy" 
<bucy-openafs-info@gloop.org> wrote:

> On Tue, Jan 18, 2005 at 02:44:35PM -0500, John S. Bucy wrote:
>> On Mon, Jan 17, 2005 at 04:09:50PM -0500, John S. Bucy wrote:
>> >
>> > I have a large (~10G) volume dump taken from a server with a failing
>> > disk.
>> >
>> > When I try to restore it, I get
>> >
>> > Restoring volume user.roboclub.folders Id 536889153 on server
>> > ALUMINUM.CLUB.CC.cmu.edu partition /vicepa ..rx_EndCall Failed VOLSER:
>> > Problems encountered in reading the dump file !
>> > Error in vos restore command.
>> > VOLSER:  Problems encountered in reading the dump file !
>> >
>> >
>> > (and I got a similar error when it finished dumping)
>> >
>> > Will dumptool help me here?  Any other ideas?
>>
>> The volume seems to have restored anyway though I'm not yet sure how
>> corrupted it may be.  What does this error mean?
>
> In VolserLog, I have
>
> Mon Jan 17 15:44:08 2005 1 Volser: WriteFile: Error reading dump file
> 119806 size=11743 nbytes=11743 (7280 of 8192); restore aborted Mon Jan 17
> 15:44:08 2005 1 Volser: ReadVnodes: IDEC inode 119806
>
> after a lot of whining about the transaction taking a long time.


Well, the transaction taking a long time is normal, and the volserver is 
very good at whining about it.  The error in the volserlog is not one I've 
seen before; I wonder if it's a problem with the volserver or the dump.

I've had a certain amount of experience patching up broken volume dumps; I 
can take a look at it if you want.  We probably should take this offline, 
though...

-- Jeff