Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.rmit.EDU.AU!news.unimelb.EDU.AU!munnari.OZ.AU!news.ecn.uoknor.edu!paladin.american.edu!gatech!news.mathworks.com!nntp.primenet.com!news.cais.net!van-bc!nntp.portal.ca!cynic.portal.ca!not-for-mail From: curt@cynic.portal.ca (Curt Sampson) Newsgroups: comp.unix.bsd.misc Subject: Re: Queuing Question Date: 26 Jun 1996 12:26:43 -0700 Organization: Internet Portal Services, Inc. Lines: 25 Message-ID: <4qs2tk$b7o@cynic.portal.ca> References: <4qn5nu$svs@pulp.ucs.ualberta.ca> <31D09F31.6DB0F89A@lambert.org> NNTP-Posting-Host: cynic.portal.ca In article <31D09F31.6DB0F89A@lambert.org>, Terry Lambert <terry@lambert.org> wrote: > >I guess one way would be to put a time limit on the job and >murder it if it goes over the limit....This is pretty useless >without checkpoint/restart technology of some kind so you >can resurrect the murdered job at some future time. Actually, you could just give it a SIGSTOP and keep a queue of jobs to unsuspend, starting them up again in they order they stopped. You will, of course, want a fair amount of swap space on your system if you do this. :-) And yes, there are probably ways around this, but it seems likely that users also have timesharing access to this system, allowing them to start up any jobs they want anyway. Even if they don't, it's most likely cheaper just to keep an eye out for the `pigs' and have a chat with the person submitting the job than it is to do the work necessary to make the thing really secure and fair. cjs -- Curt Sampson curt@portal.ca Info at http://www.portal.ca/ Internet Portal Services, Inc. Vancouver, BC (604) 257-9400 De gustibus, aut bene aut nihil.