[OpenAFS-devel] performance study

[Ken Hornstein]

>AFAIK, noone has done a detailed performance evaluation of AFS in some
>time.  I was thinking of doing such a thing for a class project. 
>Here are a few things that I've thought about looking at:
>
>1.  RX.  How does it do on the wire?  How does it compare to modern
>sunrpc+gssapi?  How does it compare to TCP for bulk-data throughput.

My boss has actually tasked me with this.

I've not compared RX against sunrpc+gssapi, but I've compared it
against TCP on higher speed networks.  Between two OC-12 connected
machines, I can get around 510-520 Mbits.  The best I can do with RX
between the same machines is around 140 Mbits (I know Hartmut has
claimed that he sees full bandwidth at Gig-E speeds, but I've never
seen that here).

--Ken