🚀 go-pugleaf

In article <26642@shemp.CS.UCLA.EDU> frazier@cs.ucla.edu (Greg Frazier) writes:
...deleted stuff

>machines experience a wide range of workloads, and cannot be
>customized for each one of them.  Any heavily-used scientific
>machine is going to experience the described situation at some
>point in its career.  Now, obviously, if you have somebody
>monitoring the machine's performance, he can "manually" prevent
>the two jobs from running simultaneously.  What would be more
>desireable, however, is for the OS to realize what is going on,
>and for _it_ to cause the jobs to run sequentially.  I suspect
>that this would require too much intelligence on the part of the
>OS, but then, what do I know - I'm just an architect! :-)

These are reasons why some folks actually like batch OS ... in the old
days we had ways of informing the system of our expected needs (cpu,
io, "cards", etc.) and this was used by some fairly clever scheduling
algorithms. As Unix machines grow up, I personally expect something
fancier than nice level to allow the savvy programmer to inform the OS
of expected needs (for these scientific codes, we should be able to
assert, long running, not very interactive .. this would enable long
timeslices but at low priority ... and perhaps only one mongo job at a
time). 

There are some "batchqueue" type products which third parties and
national labs have cobbled up.

Keith H. Bierman    |*My thoughts are my own. !! kbierman@sun.com
It's Not My Fault   |	MTS --Only my work belongs to Sun* 
I Voted for Bill &  | Advanced Languages/Floating Point Group            
Opus                | "When the going gets Weird .. the Weird turn PRO"