FNN vs GigabitEther & Myrinet

Bogdan Costescu bogdan.costescu at iwr.uni-heidelberg.de
Thu Oct 25 14:36:18 PDT 2001


On Thu, 25 Oct 2001, Mark Hahn wrote:

> I think this is a bit distorted (and seemingly reactionary).

I didn't mean it. I apollogize if it really looked like...

> ... these factors were too
> specialized to reasonably apply to zero-copy networking.
>
> for 2.5, expect a modestly modified version of the networking zero-copy stuff
> to replace the old disk-oriented stuff.

My point was that there are now 2 different mechanisms - one or maybe both
can be used with something like GAMMA. The introduction _in a stable
kernel_ of a second such interface was a mistake IMHO. Especially if, as
you say, this is the one that will stay. So people trying to implement
something to plug in the kernel are forced to change their minds if they
picked the wrong interface - which is always difficult, especially when
this is open-source and commitment is at own will. So, apart from
supporting the ever changing hardware, they have to support the ever
changing software as well ! Of course that this will discourage anybody
and that's why I thought that involvment of a company/university/other
insitution would motivate more the developers.

Sincerely,

Bogdan Costescu

IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu at IWR.Uni-Heidelberg.De






More information about the Beowulf mailing list