Xref: utzoo comp.arch:11103 comp.sys.mips:108
Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.csd.uwm.edu!gem.mps.ohio-state.edu!ginosko!uunet!mcsun!ukc!edcastle!adam
From: adam@castle.ed.ac.uk (Adam Hamilton)
Newsgroups: comp.arch,comp.sys.mips
Subject: Re: Memory utilization & inter-process contention
Message-ID: <218@castle.ed.ac.uk>
Date: 23 Aug 89 12:53:14 GMT
References: <3332@blake.acs.washington.edu> <1989Aug22.163100.25540@utzoo.uucp> <3342@blake.acs.washington.edu>
Reply-To: adam@castle.ed.ac.uk (Adam Hamilton)
Organization: Edinburgh University Computing Service
Lines: 9

In article <3342@blake.acs.washington.edu> lgy@blake.acs.washington.edu (Laurence Yaffe) writes:

	We have an ongoing discussion about scheduling of large jobs.
I suggest that a good batch system would be better at this than trying to
do everything in the kernel.  It can monitor the various requirements of
the separate jobs (store utilisation, i/o and system call patterns etc)
and tune accordingly.  What it needs from the operating system is this
information plus some more (like interactive usage).
	Summary - if it doesn't need to be in the kernel, don't put it there.