[OpenAFS-devel] Fileserver as Masters Thesis

[Niklas Edmundsson]

Hi all!

Me and a colleague are planning on finally getting our Master Thesis
done, and our current plan is to implement a new OpenAFS fileserver
since that would be a rather useful result of 20 weeks work.

We envision not touching the RX stuff, hoping that rxtcp will solve all
the various issues that's in there.

The current situation is that the INODE fileserver has it drawbacks, and
so does the NAMEI fileserver.

Ultimately one would want the performance of the INODE fileserver
combined with the portability of the NAMEI fileserver. With performance
we're talking both small-file tinkering and scaling up to next
generation network speeds and filesizes.

We see two roads to investigate:

Either do an implementation that uses an existing filesystem like the
current NAMEI fileserver, or implement an AFS-server-specific file
system.

Designing/implementing a new filesystem is a lot of work, and has a lot
of "reinventing the wheel"-feeling to it. You need good arguments to
follow this path, and it would be preferable if there is existing work to
leverage this upon. Licensing becomes an issue here.

Obviously, we welcome all kinds of input on this.

Also, any pointer to documentation, previous work etc. regarding the AFS
fileserver is welcome. We're not that clued on what subtle features
there might be that the fileserver needs to support.

Regards,
Nikke and Hannes