Re: Clustering for Linux 2.3.x?!

["Robert G. Werner" <rwerner@lx1.microbsys.com>]

Well,  I wasn't really making much of a statement about anything except just
restating some of the things I've come across in my explorations of clustering.
So I probably just fall into the catagory of stating stuff you forgot.  
I've seen a lot of wierd ideas about how clusters can be so much more powerful
than uniprocessors.  I just wanted to make sure that the constraints of
clustering were recognized befor people got too far off into blue sky ideas.

But it sounds like you know what you are talking about so I (a humble begining
enthusiast ;-) ) won't presume to tell you how to suck eggs ;-P .

Finally,  I think the message passing actually sounds conceptually easier than
trying to implement complex systems of locking (but again I may just be
speaking out of my ignorance).  I like that message passing for sure (but even
threading or shared memory systems) moves more of the burden for the program
for sheer processing to clever splitting of the problem.

Robert G. Werner
rwerner@lx1.microbsys.com
Impeach Conggress!!

While having never invented a sin, I'm trying to perfect several.

On Thu, 4 Mar 1999, Erik Elmgren wrote:

[snip]
> You claim less latency for distributed-shared-memory or that shared-memory
> programs are more latency tolerant or both? Or are you just telling me the
> stuff I forgot to include? :-)
> 
[snip]
> It has been claimed (not here) that shared-memory is superior because many
> programmers know it better than message passing.
> 
> I also forgot to say that message-passing could indeed benefit from kernel
> modification, but not from total redisign.
HOW?  Just curious what needs to be done.
[snip]

-
Linux-future: thinking about the future of the Linux kernel
Archive:      http://humbolt.nl.linux.org/lists/
Wish list:    http://users.ox.ac.uk/~mert0236/linux-future.html