Re: Remote fork() and Parallel programming

Rik van Riel (H.H.vanRiel@phys.uu.nl)
Wed, 17 Jun 1998 09:55:56 +0200 (CEST)

On Tue, 16 Jun 1998 linker@nightshade.ml.org wrote:
> On Mon, 15 Jun 1998, Emil Briggs wrote:
>
> > >The fastest TCP latency I know of is about 80,000ns. The fastest Unet
> > >(no protocols, network mapped into the process' address space) latency
> > >is about 30,000ns.
> >
> > Are you sure about the 30,000ns figure? There are several solutions
> > that claim 2-3 microsecond latency using some custom hardware and
> > userspace libraries. These include
>
> Though I can't imagine that it needs to be *that* high.. With a special
> (non-ip, non checksummed) ethernet protocall for san's and a highly
> optimized networking path much lower should be possible..

There's a special networking protocol (called flip) for
this in Amoeba. Don't know how good it is, but it might
be worth a look. Maybe ast will be friendly enough to
release the flip code under GPL... (I believe Amoeba is
free now)

> I still think my point in clear, such remote fork and migration should
> workable if the migration is infrequent enough.

Amoeba did it, I believe Sprite did it too. This proves that it
can be done, just not that it can be done efficiently...

Rik.
+-------------------------------------------------------------------+
| Linux memory management tour guide. H.H.vanRiel@phys.uu.nl |
| Scouting Vries cubscout leader. http://www.phys.uu.nl/~riel/ |
+-------------------------------------------------------------------+

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu