From: ws@xivic.bo.open.de (Wolfgang Schelongowski) Subject: Re: SCSI Performance (Yet Again) Date: Mon, 23 Aug 1993 16:37:17 MEST
jhenders@jonh.wimsey.bc.ca (John Henders) writes:
...
> SO far, it seems
> to me that the theoretical 5.7 meg/sec of the Adaptec is just that,
> theory, and only a few people get even 1 meg/sec.
The 5.7 MB/sec are the DMA transfer speed. There are other things
happening on the SCSI-Bus which have to complete before that. If you
need to know more about what's going on there, RTFM. There's a
(German) book about the IDE-Interface and SCSI-Bus (Author:
Friedhelm Schmidt, ISBN: 3-89319-597-1). As it is published by
Addison-Wesley, you can pester them to have a translation released
if there is no English book about that subject.
> Again, on my ST, there was a program that bypassed the filesystem
> completely to test raw disk i/o, and somehow they even made it
> non-destructive so you didn't have to use a clean partition to do tests.
> It seems there is no equivalent under *nix.
How about
date;dd if=/dev/sda of=/dev/null bs=131072 count=512;date
(This does only a read, you have to save and restore via dd
if you want to test write. Just be _very_ careful about _where_
you test ! Maybe you should backup the disk.)
That took 69 seconds => 950 KByte/sec for raw I/O. Under BSDI,
the numbers are 36 seconds on /dev/rsd0a => 1820 KByte/sec.
As I'm still running 0.99 pl6 the numbers for Linux may have
changed (hopefully to the better). Can somebody test it and
post the results so we have _numbers_ to compare and not
assumptions ?
My System: noname 486/33, AHA1542B, Fuji M2624FA. It makes no
difference for Linux whether I use async or sync SCSI.