home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.arch:8182 comp.sys.sun.misc:3263 comp.unix.solaris:99
- Path: sparky!uunet!darwin.sura.net!gatech!rutgers!cbmvax!jesup
- From: jesup@cbmvax.commodore.com (Randell Jesup)
- Newsgroups: comp.arch,comp.sys.sun.misc,comp.unix.solaris
- Subject: Re: Sun 600MP Benchmark Anomaly
- Message-ID: <33209@cbmvax.commodore.com>
- Date: 21 Jul 92 18:40:11 GMT
- References: <1992Jul16.122707.11010@cl.cam.ac.uk> <l6bkvtINNaf6@appserv.Eng.Sun.COM> <1992Jul17.131131.26125@cis.uab.edu> <BrKJu8.5An@acsu.buffalo.edu> <1992Jul20.131307.19497W@lumina.edb.tih.no>
- Reply-To: jesup@cbmvax.commodore.com (Randell Jesup)
- Organization: Commodore, West Chester, PA
- Lines: 22
-
- ketil@edb.tih.no (Ketil Albertsen,TIH) writes:
- >For disks with high rotation speeds (eg 5400 rpm) the arm movement time
- >is more significant than for traditional 3600 rpm disks.
- >Note also that if SSTF (Shortest Seek Time First) arm scheduling is used,
- >middle cylinders will on the average recieve better service, as the shorter
- >arm travel will push them ahead in the queue.
-
- One of the many problems when trying to improve on FCFS (First Come,
- First Served).
-
- >(Note that on most disks, the angular bit density is the same throughout
- >the surface, although the linear bit density is higher at the inner cylinder.)
-
- "Most" disks? It seems like almost all disks today are zone-recorded.
-
- --
- "Rev on the redline, you're on your own; seems like a lifetime, but soon it's
- gone..." Foreigner
- -
- Randell Jesup, Jack-of-quite-a-few-trades, Commodore Engineering.
- {uunet|rutgers}!cbmvax!jesup, jesup@cbmvax.cbm.commodore.com BIX: rjesup
- Disclaimer: Nothing I say is anything other than my personal opinion.
-