Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ames!lll-winken!uunet!mcvax!unido!pfm!nadia!space
From: space@nadia.UUCP (Lars Soltau)
Newsgroups: comp.sys.amiga
Subject: Re: Problems with Interleave on hds
Summary: Sorry, but there's no easy way to do what you want!
Keywords: SCSI interleave A2090 hddisk.device Prep
Message-ID: <362@nadia.UUCP>
Date: 23 Apr 89 16:37:39 GMT
References: <890416195745B19.AHCV@USCN.USCN.UGA.EDU>
Reply-To: space@nadia.UUCP (Lars Soltau)
Organization: Stuttgart Net System, FRG
Lines: 37

In article <890416195745B19.AHCV@USCN.USCN.UGA.EDU> bach@USCN.BITNET (BitNet Users Of Southern Tech) writes:
>Now, I've got a 2500 with A2090a controller and I want to see if I can get
>better speeds by using a different interleave.  The problem is that I'm
>getting just about the same speeds from Diskperf no matter what interleave
>I use.
Let me restate that: You are getting the same interleaving factors no matter
what interleave you tell Prep to use.

>I've been following the same procedure each time:
>1.  Change the interleave in the mountlist
>2. reboot the computer and mount the drive
>3. Prep the first partition and reboot
>4. Format the drive FFS and run diskperf V3.0
>5. Shake my head as I get the same results....
Same problem I had. I already wrote a long article about this. Maybe you
can still find it on your system, it was posted not too long ago.

As I wrote in that article, Prep doesn't give a <bleep> about the Interleave
value stated in the MountList entry for RES0: (or whatever you call it).
If you really want to change the Interleave, you have to ways:
1) Singlestep hddisk.device (with Aztec db, for example) and find where it
   sends the SCSI format command to the disk, then change the command on the
   fly and let Prep run.
2) Lowlevel-format the drive on a PC (yuck!) and mount your partition on it
   without prepping. This way you can only format the drive "QUICK" after-
   wards.

If I'm talking bullshit here, and anyone from C=A knows a simple and user-
friendly way to change interleave on an SCSI drive with the A2090(A), go
ahead and tell me, don't let me end my life in ignorance! :-) If not,
consider this an urgent plea for a fix of this bug.

BTW, is there any policy from C=A concerning updated A2090 software for us
poor A2090(no A) owners? I'd very much like to get my hands on the new
revision. I'd be willing to pay for it, too.
-- 
Lars Soltau	UUCP: ...uunet!unido!pfm!nadia!space	BIX: -- no bucks --