Products > Test Equipment
16702b SCSI ID issue
(1/2) > >>
dorkshoei:
[please don't suggest scsi2sd - I own several - but it's not question I'm asking]

I just bought another Agilent 16702b.

The original SCSI drive has failed.  It was an IBM DPSS-309170 which is a 9GB Ultra160 68pin.     There was an Agilent 16702-xxx part# sticker on the drive so I'm sure it was the original drive.   It was set to SCSI ID 6 which seems to be the OEM value.

I'm trying to install onto a  replacement Fujitsu MXJ3367SC680600X which is a 36GB Ultra160 68pin.

I found it was necessary to install the "Force narrow:  16 bit scsi" jumper (#7) on the Fujitsu  Odd verbiage as narrow is 8-bit.

More bizarrely the HP installer will only run if the drive is set to SCSI ID 0.   

If I set it to anything else (SEA in the BIOS shows I'm setting the correct ID) when the installer runs from CD I see "Scanning system for IO devices ... Querying disk device [id] reported a size of zero",   Same as reported here:  https://community.hpe.com/t5/System-Administration/Reported-a-size-of-zero-to-install-HP-UX-10-20/td-p/4506995

The drive seems to work perfectly.  I hooked it upto a PC with a Adaptec 2940UW controller and ran a block scan without issue (set to ID 6).

Anyone seen something similar?

Thanks

---

As an aside (in case it ever helps anyone else searching) when I did the install onto SCSI 0 I got a ton of errors of the form "Warning: utime of file: <path> failed.  Invalid argument (22)".  Sure enough the RTC on the board was showing 2088.  I reset it but next time it was showing the year 20135 ;-)  After taking half the machine apart I confirmed the 2325 lithium battery was reading 0.35V.      Now I just need to wait for the replacement to arrive from Amazon, uncommon size.   Hopefully this fixes the above issue.
dougg:
On an 8 bit SPI (SCSI Parallel Interface) bus the controller (HBA) usually has ID 7 and the first disk has ID 0 (also known as its LUN). So the first and only disk being at ID 6 is strange, perhaps some game being played by HP.

Anyway I know a bit about SCSI (see: https://github.com/doug-gilbert ) so if you need some help, contact me. My email is in any of those github packages.
dorkshoei:

--- Quote from: dougg on September 11, 2022, 01:59:21 am ---On an 8 bit SPI (SCSI Parallel Interface) bus the controller (HBA) usually has ID 7 and the first disk has ID 0 (also known as its LUN). So the first and only disk being at ID 6 is strange, perhaps some game being played by HP.

--- End quote ---

The maxtor cdrom is @1 and the hd @6.   Why this is I don't know.  There is no requirement in the spec for your first disc to be @0.   I may try to find another hard drive to try.   The actual cable is narrow (50 pin) but the unit came with a 50/68 pin adapter and as I said,  the 68 pin IBM drive has an Agilent OEM 16702-xxxx sticker on it.   I suspect Agilent just didn't want to update the controller and narrow drives were by this point impossible to source (product started out much earlier via HP).
gslick:
The standard 16700B/16702B SCSI CD-ROM is a Plextor drive. That's probably what you meant when you typed Maxtor.

I had at least one where the original hard drive was a 68-pin 9.1GB IBM model DNES-309170.

I have also used 68-pin 9.1GB IBM model DDRS-39130 drives to replace dead or missing 16700-series hard drives. Of those two IBM models, the DNES-309170 is slightly newer and has a data buffer that is about 4 times larger.

I have never tried using a Fujitsu hard drive in a 16700-series logic analyzer. I have no idea what might be causing the reported size error issue. Is that mainly just a curiosity, and not a functional issue if you use the drive at SCSI ID 0 instead of the usual SCSI ID 6?

If it is just a curiosity, I suppose you could either just ignore it and use the drive at SCSI ID 0, or if you are really curious, hook up a SCSI bus analyzer and try to find out what happens differently on the BUS when the drive is at ID 0 vs at ID 6.
dxl:

--- Quote from: dougg on September 11, 2022, 01:59:21 am ---On an 8 bit SPI (SCSI Parallel Interface) bus the controller (HBA) usually has ID 7 and the first disk has ID 0 (also known as its LUN). So the first and only disk being at ID 6 is strange, perhaps some game being played by HP.

--- End quote ---

On PC systems it was common to start with ID 0 for the disk. At least on HP workstations it has always been common to put the disk at ID 6. I'm not sure whether that was because of priorites - on SCSI ID 7 has the highest priority (that's why the HBA usually sits there), so it makes sense to have the SCSI Disk sitting at the second highest priority. At least i would like to have my SCSI disk have higher priority then i.e. a CDROM drive. Not sure whether that makes a difference in reality though.
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod