Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!cornell!uw-beaver!ubc-cs!alberta!calgary!xenlink!blender!bruce
From: bruce@blender.UUCP (Bruce Thompson)
Newsgroups: comp.arch
Subject: Re: Memory utilization & inter-process contention
Summary: Classic resource allocation problem
Message-ID: <88@blender.UUCP>
Date: 26 Aug 89 00:15:11 GMT
References: <3332@blake.acs.washington.edu> <1989Aug22.163100.25540@utzoo.uucp> <9aid02rf4dNn01@amdahl.uts.amdahl.com>
Organization: Some apartment in downtown Calgary
Lines: 38

In article <9aid02rf4dNn01@amdahl.uts.amdahl.com>, sbf10@uts.amdahl.com (Samuel Fuller) writes:
> 
> In article <26642@shemp.CS.UCLA.EDU> frazier@cs.ucla.edu (Greg Frazier) writes:
> >
> >The original posting was asking if there was an OS smart enough
> >to run the jobs in sequence for him.
	/* much deleted */

> The dreaded JCL requires that you tell the system how long your job
> will run, how much memory it will need, and any resources like tape
> drives that you will be using.  Your job is then classified based on
> these parameters.
	/* more deleted */

    I think that what has just been demonstrated is that memory and
processor time is a resource which must be managed by the OS on the same 
level as it manages tape units, disk drives, etc. etc. I think that it 
would be fairly widely agreed that generally it is not known either how
much memory will be required, nor how long a program will run. While in
some cases there may be advanced knowledge of the characteristics of a
program this will not be the case in general.

    I think perhaps the issue is whether or not there is a general method
which can be used to determine on an `on demand' basis how to schedule
memory and the processor(s). This is essentially a summary of what the
entire area of operating system development has been about. As I recall,
one of the reasons that operating systems have tended to move away from
pre-identification of resource requirements was due to the fact that the
requirements were not known accurately.

    In the particular case of processor time requirements, one of the
more successful strategies for processor scheduling involves multiple
priority queues. Perhaps something along these lines can be developed for
managing memory resources as well.

    Certainly, it occurs to me that there is no simple solution.

	Bruce Thompson