openmosix and clustering hardware

googlinggoogler_at_hotmail.com
Date: 02/28/05 

Date: 28 Feb 2005 00:53:28 -0800

Hiya,

Ive produced a program thats quite processor intensive, and I was
considering re-writing parts of it to fork some of the processes. I
think I know how to do this, although if not i'll just hit the libary
for a few hours.

Anyway at present im running this on a 3GHz x86, with 256mb of RAM -
with my limited knowledge of clustering and my disunderstanding of what
to search for on google, I believe that my current program using
standard C without threads and forking of processes clogs possible
pipelines(is that the correct word?) becuase of the massive amount of
data im processing.

I was woundering whether the building of a simplpe cluster with older
computer namely 3 or 4 800mhz with 256mb of ram would aid processing?

I know that in total it doesnt add up to more, but I was woundering
whether the distribution of processes would indeed help "lighten" the
load, and cause the process to run faster providing the program is
re-written to fork the processes.

Is this idea simple called distribution, and how would on go about
calculating the effectivness of going down this root before imbarking
on the implmentation and constructing

Cheers

Dave

Relevant Pages