[ClassicCMP] Re: Disk size limits in VMS 5.4-1?

From: Jim Strickland <jim_at_calico.litterbox.com>
Date: Sat Feb 5 15:26:57 2000

I believe VMS 5.4 is early enough that it doesn't like non-dec drives. I was
unable to get my seagate drive to show up at all until I upgraded to VMS 7.
whatever is on the hobbiest cdrom v2. The older versions' scsi driver was
finicky. There's a workaround, but I've forgotten what it is, and I'm sure
someone will post it.



>
> I have a very perplexing problem. What I want to do is mount a Seagate
> ST410800N on a VMS 5.4-1 system.
>
> The hardware is a VAX 4000/300 with an Emulex UC-08-III controller half of
> which is in MSCP mode. (the other half is talking to the tape drive) There
> are two existing 1GB (Imprimis) drives on this system.
>
> First attempt:
>
> Put the drive into the SCSI chain, terminate it correctly, drives are
> targets 0, 1, and 2. The new drive is now target 0, the others are moved to
> 1 and 2. I reset the VAX and get into the Emulex firmware (rev 'M' (changed
> to 'R' later, see below)) and try to autoconfigure the drives. The two 1GB
> drives show up like they should, the 9GB drive shows up as a 500MB drive.
> Weird.
>
> Second attempt:
> Replace the firmware in the Emulex with version 'R'. Attempt to
> reconfigure, same problem as #1.
>
> Third attempt:
> Attempt to manually specify the geometry and notice that the Seagate is
> reporting it has 4925 cylinders but the prompt says the number has to be
> between 20 and 4095. I enter a number smaller than 4095 and say 'OK'. Now
> the drive shows up as having 14x10E6 blocks! Ok so now we're getting
> somewhere but I forgot to configure the other two.
>
> Fourth Attempt:
> Recompute "Fake" heads/sectors numbers so that by using 4094 cylinders I
> get close to the drives 17,845,431 blocks. Configure the two 1GB drives
> with all the defaults. Seagate drive shows up as a 1/2 G drive again. On a
> whim I subtract the number of blocks shown from what it should be, the
> result is in the 16x10E6 range, aha! I say.
>
> Fifth Attempt:
> Recompute cylinder/sector/head numbers so that the total will be less than
> 2^24-1 blocks. This works and I end up with a controller that thinks the
> drive is about 8GB in size (I could live with this for now...). So we go
> all the way up to VMS.
>
> In VMS I type 'INIT DUA0: TEST' and it fails with "%INIT-F-CLUSTER
> unsuitable cluster factor", I drill down through the help files but there
> doesn't seem to be a comment on this. It talks about VAX clusters and
> CLUSTER_SIZE which I tried setting to higher numbers (like 10 or 16). But
> nothing has worked so far.
>
> My questions then are:
> 1) Is there a way to get the Emulex controller to recognize more that 2^24
> blocks on a SCSI drive?
> 2) Does VMS have a volume size limit in 5.4-1 that I need to know about?
> 3) If I "split" the volume on the Emulex into two logical drives, can the sum
> of the logical drive sizes be greater than 8GB?
>
> Curious minds want to know!
> --Chuck (Who is enjoying VMS for the time being ... ;-))
>
>
>


-- 
Jim Strickland
jim_at_DIESPAMMERSCUMcalico.litterbox.com
-----------------------------------------------------------------------
                              BeOS Powered!
-----------------------------------------------------------------------
Received on Sat Feb 05 2000 - 15:26:57 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:32:52 BST