🚀 go-pugleaf

: > 
: > As another example, some grad students have been running large
: > processes across many of the same machines for 24/7 for at least a
: > week ("src*******").  Who knows what they are doing.

: This is really obnoxious since there are RESEARCH machines in the CS sub
: domain that are there specifically for graduate students to do there
: work on!!  We poor undergrads can't even LOG ON to these machines, much
: less use up all of the available resources.


Well, it was only a matter of time before a grad student jumped into this
conversation...

I saw the process that Tony was talking about -- and it seems pretty
bizarre.  Before posting messages to a newsgroup with complaints, it is
far more effective to just ask the user (in this case, subin) if he
knows he has a process running (and hogging) these machines.  If he
doesn't kill the job or explain what is going on then you should submit
a trouble report and get the Systems people involved.  Complaining
on a newsgroup can make more people aware of the problem, but it
doesn't necessarily mean anything will be done to solve the problem.

In the defense of the graduate students, the vast majority of graduate 
students do NOT run processes on the general machines because there are 
faster machines that are SUPPOSED to be used for cycle crunching.  This
post is kind of amusing to me because we often hear the same complaints
about the so-called research machines.  I also looked at the load on
the "bogged down" machines, and they aren't nearly as bad as some of
our machines.  Trust me, the research machines are usually not sitting 
idle.  Unfortunately, some people (grad and undergrad) have to run their 
programs on the HP's -- we don't have any "research" machines that are 
HP's.  Since not all code can be ported to Sun's, I guess that means
we will all just have to share the HP's...

--Soraya
-- 
-Soraya
---------------------------------
rana@cs.colostate.edu
http://www.cs.colostate.edu/~rana