home *** CD-ROM | disk | FTP | other *** search
- Sender: Postmaster@iecc.cambridge.ma.us
- Newsgroups: comp.arch
- Path: sparky!uunet!math.fu-berlin.de!ira.uka.de!Germany.EU.net!gmd.de!newsserver.jvnc.net!louie!udel!bogus.sura.net!howland.reston.ans.net!spool.mu.edu!sdd.hp.com!think.com!spdcc!iecc!mailgateway
- Subject: Re: HOw many PC's make an Amdahl mainframe
- References: <1993Jan26.225943.21955@enterprise.rdd.lmsc.lockheed.com>
- Organization: I.E.C.C.
- Date: 27 Jan 93 16:50:43 EST (Wed)
- From: johnl@iecc.cambridge.ma.us (John R. Levine)
- Message-ID: <9301271650.AA13180@iecc.cambridge.ma.us>
- Lines: 14
-
- >[People still use mainframes for very large databases; PCs and file or
- database servers will squeeze that out]
-
- Maybe and maybe not. I've seen plausible claims that when you're handing
- data bases, 10 systems of capacity N are no substitute for one system of
- capacity 10N, mostly because the slower systems hold each lock 10 times as
- long so you get considerably more lock contention.
-
- Is there any reason to think this isn't true? I realize that for slowly
- changing data, e.g. schedules and fares in an airline system, you can
- replicate and partition the data, but they do that on mainframes, too.
-
- Regards,
- John Levine, johnl@iecc.cambridge.ma.us, {spdcc|ima|world}!iecc!johnl
-