[OpenAFS] fine-grained incrementals?

Jeffrey Hutzelman jhutz@cmu.edu
Thu, 10 Feb 2005 11:37:30 -0500


On Thursday, February 10, 2005 12:03:09 PM +0100 Frank Burkhardt 
<fbo2@gmx.net> wrote:

> Hi,
>
> On Tue, Feb 01, 2005 at 03:15:37PM -0500, Jeffrey Hutzelman wrote:
>
> [snip]
>
>> As far as processing volume dumps is concerned, take a look at my
>> dumpscan  tools (currently in
>> /afs/cs.cmu.edu/project/systems-jhutz/dumpscan).  They  provide
>> libraries for parsing and emitting volume dumps, and a few tools as
>> well.
>
> I was curious so I tried:
>
> $ dump root.cell -file root.cell.dump
> $ afsdump_dirlist root.cell.dump
>   VNode      Uniqifier   Name
>   ========== ==========  ==============================
> afsdump_dirlist: Bad magic number in AFS volume dump Invalid page tag
> (41235) in page 0 *** FAILED: Bad magic number in AFS volume dump

afsdump_dirlist operates on files that contain a single AFS directory, as 
would be found on the fileserver's disk or in a cache chunk.  It won't work 
on volume dump files.



> $ afsdump_extract root.cell.dump
> chdir (null) failed: Bad address

ISTR the argument parsing in afsdump_extract is not too great.  You may get 
better results if you actually give it the name of a directory to extract 
into.

If you just want to look at what's in the dump, you want afsdump_scan. 
Don't just run it -- look at the help; there are lots of options for 
controlling how much output it generates.

-- Jeffrey T. Hutzelman (N3NHS) <jhutz+@cmu.edu>
   Sr. Research Systems Programmer
   School of Computer Science - Research Computing Facility
   Carnegie Mellon University - Pittsburgh, PA