lkml.org 
[lkml]   [1999]   [Mar]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: MOSIX and kernel mods.
Kamran Karimi writes:
>
> On Fri, 5 Mar 1999, Richard Gooch wrote:
> > > I agree completely. There will be some overhead when using systems
> > > like DIPC. That is because they are taking over some of the work
> > > that should be done by the programmer if he uses systems like PVM or
> > > MPI. Actually, this is true even in the case of PVM and MPI, as they
> > > rely on TCP sockets. I wonder if the best solution in your opinion
> > > is to program the networking adaptor directly?
> >
> > Yep, bugger the network stack, I write to the NIC registers directly
> > and grab the interrupts too. I use a dedicated NIC and network for all
> > my distributed applications. Nothing beats fondling the bare metal!!!
> > I use RT-Linux too to squeeze out the last drop of performance: I'm
> > aiming for sub-microsecond interrupt latency!
>
> Thanks. You just proved my whole point in my previous postings.

Groan. Ding, ding, ding. That was sarcasm.

> > A message passing API for an SMP machine is the wrong abstraction,
> > because it also prevents you making best use of the machine (although
> > it's not as foolish as threading/SHM for a distributed machine).
>
> The real foolish thing is to think that what is suitable for one
> person is suitable for all others, and what is not suitable for one
> person is not suitable for others.

As I said before, if it's just for you and your friends, and you
understand the pitfalls, then fine (of course, then patches to the
official kernel aren't appropriate).

If you're providing this for others, then I think you've doing them a
disservice, because you're giving something that looks really nice,
but has too many hidden traps.

> > What I object to is this idea that you can produce a single, unified
> > abstraction solve all problems. It just doesn't work this way.
>
> This is what I have been saying. I never claimed that DIPC is the
> solution to all distributed programming needs, but it covers many
> areas.

It is my view that DSM is not an appropriate model for most classes of
real applications. DIPC minus the DSM bit is a different story. But
then we're talking about MPI. I've got no problems with MPI (not that
I've used it, since I've got my own API).

Regards,

Richard....

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:50    [W:0.048 / U:0.656 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site