init disk drive

[scc_4]

Hello,
I have a Seagate ST31230N that I like to used it on my MicroVAX model 80. I think this disk is essentially a RZ26N but with the SCSI modepage bits set differently. Does anyone know what the modepage settings are for the ST31230N to make it into a RZ26N?


Thanks !
Scc

[Steven Schweda]

> MicroVAX model 80

MicroVAX 3100 model 80?

Have you tried using the ST31230N? Does it
not work? Are you more interested in a drive
which says that it's an RZ26N, or in a drive
which works?

I think that you're working too hard.

[Willem Grooters]

I'm not sure whether it would cause havoc (to VAX, Disk or both), but what if you just plug it in (after youy set the LUN coreect, that is)? If it works, I wouldn't mind the |D|I|G|I|T|A|L| naming.

FWIK, it is a setting in some of the drive's hardware or in the manufacturer's sectors on disk that you cannot access.

[Hoff]

Hello and welcome to the ITRC forums.

Reading much into the question posed, you're probably running an OpenVMS VAX version prior to V6.2, and you're probably slamming into the required setting for ARRE and AWRE. See the OpenVMS FAQ http://www.hoffmanlabs.com/vmsfaq for some related details. You need to figure out how to upgrade to V6.2 (or later), or how to reset the drive itself to clear these bits.

Alternatively, you are attempting to shadow the drive with an existing RZ26N drive, and there are differences between the two drives. OpenVMS VAX doesn't provide dissimilar device shadowing (DDS) and such, so you're left to acquire a couple of disks that are identical, transfer the existing bits over, and shadow those disks.

Short of figuring out how to load variant firmware into a drive (assuming you have access to variant and compatible firmware) or patching the existing firmware to alter its identity, I'd not expect there is a particular way to reset the drive identification string.

If I've not guessed correctly here, could you let me know -- and provide a little more detail and background on what you are up to?

Stephen Hoffman
HoffmanLabs LLC

[scc_4]

Thank you for all those who replied!

To further clarify my original question, my hardware is MicroVAX 3100 model 80. I am indeed running a very old version of VMS, 5.5.

Really I don't care about the string ID of the drive as long as I can init the disk and use it.

I've actually changed the ARRE and AWRE bits on a Seagate ST31051N and made it work on my MicroVAX. I changed these bits on a PC with an Adaptec 2940UW PCI card using a utility that I found on the interent and then move the drive to the MicroVAX. I've done the same on this ST31230N but it would not work. I just thought that there may be some other bits that I need to alter to make this drive work in my environment.

May be this drive will only work under OpenVMS 6.2 and later.

If anyone has further insight into this, please post a reply.

Thanks and regards,

SCC

[scc_4]

I forgot to mention that I am not trying to do anything fancy with this drive. I just want to run it as an "ordinary" disk drive.

By the way, the drive itself is good. I can operate the drive under Windows XP with the 2940UW (without changing the AWRE and ARRE bits, of course) without any problems.

Regards,

SCC

[Steven Schweda]

> as long as I can init the disk and use it.

My psychic powere are failing me today. What
happens when you try? Can you substitute a
different drive and get success, or do we
still need to worry about cables,
termination, and so on? Does the problem
disk appear in a console-mode SHOW DEVICE
listing?

> I just want to run it as an "ordinary"
> disk drive.

What do you mean by "ordinary"? Non-system
disk, or what?

Over the years, I've run my VAXstation 3100
model 30/38 antique collection using all
kinds of Seagate drives with various VMS
versions (normally V5,5-2 and later, I'll
admit) without serious problems. I currently
boot from an ST11200N, with an ST19171 data
disk (VMS V7.3, however). Other than an
occasional long ID string which confused the
old firmware, I've never had a problem.
Unfortunately for comparison purposes, all my
_really_ old VMS versions are on Q-bus VAXes
with MSCP-emulating adapters, so success
there doesn't say much about what an old
DKDRIVER might do,

[Hoff]

If you want to work with arbitrary SCSI devices such as this Seagate Hawk drive, the SCSI device drivers present with V6.2 and later are somewhat a better choice, and OpenVMS Alpha V6.2, V7.x or V8.x are yet better choices.

Older hardware isn't as tolerant of the differences among various old SCSI widgets.

If you want to pursue this Hawk (2?), start digging around in the error log. And run a differences between a working SCSI drive and this one. You're doing your own device integration and testing here, after all. (Folks have been having issues with this drive and OpenVMS for some years now, too. I'm not aware of a published solution.)

For a related tool, also see rzdisk:
http://h71000.www7.hp.com/freeware/freeware40/rzdisk/

Alternatively, real DEC disks are available and are often available cheaply. I'd suggest acquiring one. Or get a whole BA350 or BA353-class StorageWorks shelf filled with old(er) drives. (There are some counterfeit SCSI drives out on the market, so choose a reputable vendor if you select this course.)

[scc_4]

Thanks for the replies from Steven and Hoff!

The drive was shown when SHOW DEV was issued. But when I tried to INIT the disk, it would give me some kind of error (I cannot remember the exact error message) and therefore would not init the drive. I just wanted to use the drive as a data drive. I thought changing the ARRE and AWRE bits was the only thing I needed to do to make it work. Apparently there are more subtle things in integrating a SCSI disk to VMS than meet the eyes.

I suppose that the recommendation from Hoff is sound as there are used DEC drives out there that I can acquire. I guess it is my stubborness in thinking that this drive should work with an older VMS release.

Regards,

SCC

[Hoff]

*** Apparently there are more subtle things in integrating a SCSI disk to VMS than meet the eyes. ***

You have *no* idea.

It's a general issue with SCSI and IDE and USB devices, not just OpenVMS. From watching the bus traffic, I'm often surprised some of the devices I've encountered work *at* *all*.

Ancient SCSI drivers and ancient software make this all the more difficult, too. In this case and barring updates, this driver stack and this software dates c. 1991 / 1992. V6.2 (c. 1995) was the first driver tier where the drivers were set to work with and to adapt to whatever was connected, rather than with specific devices and anything that matched.

As the old joke goes: none of the S's found in SCSI are "Standard". Older SCSI is a far more rocky path than is newer SCSI.