Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!mcsun!ukc!edcastle!aiai!richard From: richard@aiai.ed.ac.uk (Richard Tobin) Newsgroups: comp.arch Subject: Re: ~8-job "knee" in response curves on Suns (was Re: IBM RS6000) Message-ID: <3987@skye.ed.ac.uk> Date: 22 Jan 91 13:12:07 GMT References: <1991Jan10.214122.9506@news.arc.nasa.gov> <5257@auspex.auspex.com> <3956@skye.ed.ac.uk> <3964@skye.ed.ac.uk> <4972@lure.latrobe.edu.au> Reply-To: richard@aiai.UUCP (Richard Tobin) Organization: AIAI, University of Edinburgh, Scotland Lines: 21 In article <4972@lure.latrobe.edu.au> CCHD@lure.latrobe.edu.au (Huw Davies - La Trobe University Computer Centre) writes: >PS I doubt the answers for > 100 processes.... The code was hacked up on the spot and lacks any checks for pipe() and fork() failing. Once the process or file table fills up, results will be nonsense. >Procs: 1 - 306 switches / second The one process case is of course meaningless, and the code doesn't check for it. As has been pointed out, the program is as much a test of piping as of switching between processes. Remember that its purpose was to find a knee in the results, not make absolute measurements. -- Richard -- Richard Tobin, JANET: R.Tobin@uk.ac.ed AI Applications Institute, ARPA: R.Tobin%uk.ac.ed@nsfnet-relay.ac.uk Edinburgh University. UUCP: ...!ukc!ed.ac.uk!R.Tobin