denbeste@bbn.com (Steven Den Beste) (06/20/89)
I've got a Starboard II loaded with RAM, and I just picked up a SCSI module for it for my two HD's. One of them is an ST277N. I've got it working, and by using FFS and by extensive playing with interleave factors I've managed to get the fastest transfer speed reported by "diskperfa" to the not-very-impressive value of 70K. I have yet to try their speed-up option that disables checking, but I don't anticipate it will buy me much. [Believe it or not, I get my best performance with an interleave of 10.] Several people here on the net have claimed they got transfer speeds above 200K with the ST277N. The only thing I can think of that might yet be my problem is that I might have an old version of the driver. (Or maybe a new, broken one.) I dumped it using "type opt h" and near the beginning is the following: MAS-Drive_20 *0.5 JD version 0.32 Could someone out there who is getting great performance please let me know what version of the driver they are using? I would greatly appreciate it. Steven C. Den Beste, BBN Communications Corp., Cambridge MA denbeste@bbn.com(ARPA/CSNET/UUCP) harvard!bbn.com!denbeste(UUCP)
rmk@frog.UUCP (Rick Kelly) (06/21/89)
In article <41688@bbn.COM> denbeste@BBN.COM (Steven Den Beste) writes: >I've got a Starboard II loaded with RAM, and I just picked up a SCSI module for >it for my two HD's. One of them is an ST277N. > >I've got it working, and by using FFS and by extensive playing with interleave >factors I've managed to get the fastest transfer speed reported by "diskperfa" >to the not-very-impressive value of 70K. I have yet to try their speed-up >option that disables checking, but I don't anticipate it will buy me much. >[Believe it or not, I get my best performance with an interleave of 10.] > >Several people here on the net have claimed they got transfer speeds above 200K >with the ST277N. The only thing I can think of that might yet be my problem is >that I might have an old version of the driver. (Or maybe a new, broken one.) STUFF DELETED ... > >Steven C. Den Beste, BBN Communications Corp., Cambridge MA >denbeste@bbn.com(ARPA/CSNET/UUCP) harvard!bbn.com!denbeste(UUCP) The SCSI command set does not actually contain any modeselect commands for format interleave. When you do a low level format on a SCSI drive you are really telling the SCSI driver to tell the hard disk to execute its own format program. Seagate has taken some of its older ST-506 drives and replaced the controller board on the bottom of the drive with a SCSI controller. Transfer rate is almost always affected by the SCSI host controller on the system bus and the driver software provided. The Starboard has been around for a while, and it is not the most efficient (if you want to really slow down try a Tiny Tiger). The interleave commands that you send to the drive are affecting the Starboard and the software driver, but the hard disk will always have an actual interleave of 1. Rick Kelly Test Engineering Charles River Data Systems 983 Concord St. Framingham, Massachusetts 01701 508-626-1011
silver@cup.portal.com (Jim B Howard) (06/22/89)
That is EXACTLY the same problem I am having. The 70k per second
sounds about right. Microbotics CLAIMS I should be getting
150-200k per second. Nobody can give me a solution, and Im
getting ready to chunk this thing in the "spare parts" droor.
Ive tried FFS, different interleaves, fastmode, and even a whole
new computer system(a neighbors) but I still get 70k per second.
HELP!!!!!mp1u+@andrew.cmu.edu (Michael Portuesi) (06/24/89)
> Excerpts from ext.nn.comp.sys.amiga: 20-Jun-89 Version of driver for > Starb.. Steven Den Beste@bbn.com (1138) > I've got it working, and by using FFS and by extensive playing with > interleave > factors I've managed to get the fastest transfer speed reported by > "diskperfa" > to the not-very-impressive value of 70K. I have yet to try their speed-up > option that disables checking, but I don't anticipate it will buy me > much. Do try it -- my experience is that it about doubles the speed of I/O to and from the drive. --M -- Michael Portuesi * Carnegie Mellon University INTERNET: mp1u+@andrew.cmu.edu * BITNET: mp1u+@andrew UUCP: ...harvard!andrew.cmu.edu!mp1u+ MAIL: Carnegie Mellon University, P.O. Box 259, Pittsburgh, PA 15213