Greetings all and many thanks to those who took the time to reply!
The response was positive: the faster disks will work on the slower
controller (my particular case involved the yet to ship 10MB/sec SUN
drives for SPARC 10s running on the older SS2s (a 5MB/sec ctrlr.))
SCSI devices will negotiate a mutually acceptable transfer rate
amongst themselves. Some of the responses were based upon knowledge
of the SCSI bus and others were based upon actual running
configurations.
Others pointed out that the higher transfer rate is somewhat
misleading as seek time and intelligent controllers make more of a
difference. In particular Kevin pointed to the Winter 92 USENIX
proceedings for an "excellent discussion of this aspect of
performance" ; see "How and Why SCSI is Better Than IPI for NFS -
Bruce Nelson & Yu-Ping Cheng of Auspex Systems - p253"; thanks Kevin!
Again many thanks to the sun-managers family and in particular:
"Anthony A. Datri" <datri@concave.convex.com>
canuck@rice.edu (Mike Pearlman)
John DiMarco <jdd@db.toronto.edu>
zeke@mpl.ucsd.edu (Rob Scott)
Mike Raffety <miker@sbcoc.com>
mks!andy@watserv1.uwaterloo.ca (Andy Toy)
Petri Ojala <ojala@funet.fi>
birger@vest.sdata.no (Birger A. Wathne)
mcneill@devon.co.uk (Keith McNeill)
Wilson N G <noel@essex.ac.uk>
ukw!kevin (Kevin Sheehan {Consulting Poster Child})
cyerkes@jpmorgan.com (Chuck Yerkes)
have a better one,
david
________ __________________________ __
David Holmes McGILL UNIVERSITY
UNIX Support Group - Computing Centre Montreal, CANADA
email/NeXT email delphys@CC.McGill.CA voice 514-398-3716
delphys@MCGILL1.BITNET FAX 514-398-6876
-----------------------------------------------------------------
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:44 CDT