[ale] mosix clusters?

James P. Kinney III jkinney at localnetsolutions.com
Sun Jun 23 22:38:05 EDT 2002


If the plan is to get some sort of execution speed-up from a clustered
environment, the code must be written to support parallelization. As was
mentioned earlier, if there is a HUGE number of INDEPENDENT calculations
that can be made, that would benefit from a cluster. It won't be quite
an N-way speed up, but it will be somewhat linear (slope less that 1 due
to network latencies, code efficiencies, etc). But the usefulness of
Mosix clustering is limited to network capability and some internal
scaling issues with Mosix itself. 3 years ago when I was tinkering with
it, node counts higher than 50 were a problem. The node management code
was improved and that number is in the low hundreds now. But it is still
a _very_ finite limit. 

For my simulation code, I was forcing a synchronization step. Every x
number of iterations through the loops was the equivalent of a certain
time elapse in the simulated environment. As I understood the physics,
my "boundaries" had to "react" with in that time frame to have the
correct physics happening. So, crunch numbers for one iteration (the
compute intensive part), calculate new boundary conditions, pass
boundary conditions to "adjacent cell" nodes, normalize boundary
conditions to reflect the new conditions just received, drop state to
temp file, crunch interaction numbers again.

This was lots of fun until I realized I had commented out the
temperature routine. So I turned it back on and began a simulation of a
solid (pure lithium) being heated to the melting point. The code design
did not include molecules leaving the nodes calculation space!

On Sun, 2002-06-23 at 21:04, Christopher Fowler wrote:
> My point is that some clusters like beowulf require an API to take
> advantage of them.  You have to write your code to use the cluster. 
> Otherwise you have wasted your money.  I do not know much about a Mosix
> cluster.  So I'm not sure if it requires message passing or some other
> form of communications to get heavy computations done. SETI is an
> example that would benefit from a Mosix cluster.
> 
> 
> On Sun, 2002-06-23 at 18:49, Stephen Turner wrote:
> > i dont think it has an api, i believe i read that in the doc's, however
> > some clusters have the extra node(s)  as mirror images of the rest, so
> > they all contain the exact image hard disk, mosix says it doesnt have to
> > have the same hardware or size disks, thus i ask, does each node get
> > mirrord and all of them have the same hard disk image? or does all the
> > hard disks become one very large disk? and if so what happens to the data
> > when one node crashes? just gets lost?
> > 
> > __________________________________________________
> > Do You Yahoo!?
> > Yahoo! - Official partner of 2002 FIFA World Cup
> > http://fifaworldcup.yahoo.com
> > 
> > 
> 
> 
> 
> ---
> This message has been sent through the ALE general discussion list.
> See http://www.ale.org/mailing-lists.shtml for more info. Problems should be 
> sent to listmaster at ale dot org.
-- 
James P. Kinney III   \Changing the mobile computing world/
President and CEO      \          one Linux user         /
Local Net Solutions,LLC \           at a time.          /
770-493-8244             \.___________________________./

GPG ID: 829C6CA7 James P. Kinney III (M.S. Physics)
<jkinney at localnetsolutions.com>
Fingerprint = 3C9E 6366 54FC A3FE BA4D 0659 6190 ADC3 829C 6CA7 




---
This message has been sent through the ALE general discussion list.
See http://www.ale.org/mailing-lists.shtml for more info. Problems should be 
sent to listmaster at ale dot org.






More information about the Ale mailing list