My original question follows the summary, thanks to Tom Crummey who came
up with the correct answer, also thanks to Bismark Espinoza and Michael
Salehi who suggested other avenues to pursue.
Tom Crummey wrote:
> Try turning off tagged queueing in the kernel. I seem to remember that
> certain firmware revisions of the Baracuda didn't work with SUNs
> properly.
I did a Deja News search based on Tom's suggestion and I found a
pointer from Andrew Gabriel:
Andrew Gabriel wrote:
> You disable tagged command queueing by adding a setting to
> /etc/system and rebooting. I can't remember exactly which bit
> does it (notes not here), but I have the following in mine:
>
> set scsi_options=0x178
>
> (IIRC, one of these also turns off wide negotiation, which is
> probably not want you want if you have any wide scsi on your
> system.)
I still need to figure out how to enable tagged command queueing
on the drive, if possible, and would also like to find out what
all the bits in the set scsi_options command do, but now at least
I can run format with the -e switch so I can view/edit the drive's
mode pages.
While searching for detailed information on SCSI mode pages, I
found a DOS utility from Seagate that has some good information
in its doc files about mode pages:
ftp://ftp.seagate.com/techsuppt/seagate_utils/aspiid15.zip
> ASPI-ID v1.5 Size = 49K DOS
> A Mode Sense and Inquiry utility for SCSI interface disc drives.
> Detailed information about each of the mode pages is given along with a
> SCSI Inquiry report. This program operates through the standard ASPI
> device driver supplied by the SCSI host adapter manufacturer. This
> program can tell you more about your SCSI drive and its current
> operating parameters. Optional save of Mode Sense data to a file for
> later reference. Very detailed documentation provided. (23-May-95)
Cheers,
Mark
-- Mark C Olson | ADC Telecommunications |telephone: +1 612 936 8008 PO Box 1101, MS 125 | FAX: +1 612 936 8333 Minneapolis MN 55440-1101 | mailto:mark_olson@adc.com Mark Olson wrote: > > I have eight Seagate Barracuda ST15150N (50 pin single ended 10MB/s > SCSI) drives that were removed from an Auspex file server. I would like > to use them in external enclosures attached to Sparc 5 workstations. > > The problem is, when I invoke the Solaris (2.6) format command, it > sits here forever: > > dewey# format -d c0t2d0 -e > Searching for disks... > > And eventually after about 30 seconds, the following message starts > repeating endlessly on the console: > > WARNING: /iommu@0,10000000/sbus@0,10001000/espdma@5,8400000/esp@5,8800000/sd@2,0 (sd2): > device busy too long > > I cannot kill -KILL format, and have to reboot the system. > > I am certain I have the proper jumpers, termination and cabling[1]. > I have done a boot -r. The boot drive is ID 3 as usual, and the > ST15150N is ID 2. I have checked out the setup with two other > non-Auspex drives and they work fine, so I am confident that it > is not a physical configuration problem. > > I am more or less focusing on this being some setting in a mode page, > but since I cannot get format to even start up in the -e(xpert) mode > I cannot poke around in the mode pages, even if I knew what page(s) > to edit. > > The drive reports the following from scsiinfo: > > 2 dewey:/home/olsonm=> scsiinfo -p > esp0: sd2,0 tgt 2 lun 0: > Synchronous(10.0MB/sec) Clean CanReconnect > Non-removable Disk: SEAGATE ST15150N 4611 [SLT] > esp0: sd3,0 tgt 3 lun 0: > Synchronous(10.0MB/sec) Clean CanReconnect > Non-removable Disk: CONNER CFP2107E 2.14GB 1420 [ST] > > I do get a slightly different string containing the AUSPEX name from > probe-scsi at the ok prompt, so it is possible that they have special > firmware installed. Since the drives do work fine on a PC, I suspect > that they could be made to work fine on a Sun if only I had the secret > decoder ring... > > TIA for any suggestions, > Mark > -- > Mark C Olson | > ADC Telecommunications |telephone: +1 612 936 8008 > PO Box 1101, MS 125 | FAX: +1 612 936 8333 > Minneapolis MN 55440-1101 | mailto:mark_olson@adc.com > > [1] I know the drives themselves actually work, as I have low-level > formatted one via the SCSI controller's BIOS in a PC, and fdisked, > copied data, etc. and booted from one in a PC. I believe I have > tried nearly every sane configuration including enabling/disabling > parity, termpwr on/off, etc. I have of course also tried format > without the -d and -e options. I returned the drive to the VAR who > sold it, they pronounced it fit and said it should work just fine > on a Sun, but I suspect they did not try it on a Sparc Solaris 2.6 > machine.
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:14:03 CDT