[OpenAFS-devel] Re: IPv6 support

[omalleys@msu.edu]

Quoting Andrew Deason <adeason@sinenomine.net>:

> On Sun, 13 Feb 2011 10:29:15 +0530
> ambarisha b <b.ambarisha@gmail.com> wrote:
>
>> Yeah, I can understand.I hadn't considered the issue of
>> standardisation.May be it is best that we wait a bit more , like you
>> say.
>
> There is still quite a bit of progress that can be made before
> standardization, though, as Simon has noted. Pretty much all network
> communication in AFS goes over a protocol called RX, and you could add
> IPv6 support to the RX library without needing to standardize
> wire-protocol changes.

Maybe some documentation in the code where the changes need to occur  
would be useful? If it is, then it would also be useful to have a  
standardized status attached to the comment so you can search for it,  
ie // ipv6 support needed, //ipv6 support done.


> I'd think the first point where you see actual useful functionality in
> AFS is when you can communicate with an AFS database server over IPv6.
> It is possible to get to that point without any standardization changes
> (if you use SRV records to locate the dbservers), though I don't really
> have an idea on whether that is reasonably achievable over a summer.

I don't either but I think that is the next place I would start also.
With a single client and single server platform and probably on the  
same machine or a couple of vms on the same machine.

>> Presently,I am going through the documentation and the bug tracker,
>> hoping to make some contribution.Do you guys suggest any particular
>> project , you want done, for the summer ?
>
> A project that has been mentioned before that I would very much like to
> see is a userspace NFS/AFS translator. I don't have a pointer to more
> details on the project on-hand, but it was basically taking some NFS
> server framework, and plugging it in to libuafs (a library for talking
> to AFS).

This would be handy.

64-bit native server binaries?
I assume most of this is going through the code and checking types and  
code clean up, fixing alignment errors, etc. And -probably- one of the  
better places to do this is and I know I will get killed for saying  
this is sparc/solaris, linux/ppc or arm, (IBM power arch?) since they  
don't have hardware support for alignment errors. :)  Suns compiler  
can be overly stinky about these on sparc. I haven't used IBMs but I  
am assuming it is similar on ppc. gcc not necessarily.

A test suite functions?
There are actually quite a few real jobs I have seen posted for this  
for people with experience lately. Given the nature of this project it  
would make sense to be able to give some support for testing it.

> That said, I thought we were still unsure whether OpenAFS was doing GSoC
> this year...?

Google supposedly expanded their GSoC money.