[comp.unix.ultrix] Installing ULTRIX WS 2.0 on a DS3100

bryan@GOMEZ.phys.virginia.edu (10/11/90)

Hi folks,

	We recently purchased a DECstation 3100, and soon thereafter
corrupted its system disk while fiddling with the swap space.  We have a
complete backup, so it shouldn't be a big problem to go back to the
configuration we had before.  We'd planned to do this by booting a 
standalone system from tape, then either re-installing the original 
ULTRIX distribution and restoring from the backup tape, or using the
standalone 'restore' to just restore the root directory from the 
backup tape.
	Unfortunately, the only standalone ULTRIX tape we have available
is an ancient version 2.0 (specifically, ULTRIX WS V2.0 SUPPORTED, 
AQ-MM11A-BE), and it doesn't seem to work.  The standalone system boots,
and we can perform the patch semi-described in the ULTRIX release notes
for version 3.1D/UWS 2.2D (page A-3), in which we edit install.1 and
change 'root_D=$2' to 'root_D=RD54'.  (I'm told this is necessary for
RZ56 disks because the installation software doesn't have partition 
tables for the RZ56.)  Re-starting the installation procedure, and
choosing the item 'BASIC INSTALLATION' from the resulting menu, we see
a message indicating that 'mkfs' has successfully created the root 
file system on partition 'a' of the disk, then we get a string of error
messages like the following:

	TK50 unit 5: receive diagnostics command failed.
	TK50 unit# 5: offline
	TK50 unit# 5: offline
	read after eom failed: no such device or address.
	. 
	.
	.
	(repeats ad infinitum)

This seems to result from the installation procedure's calling 'restore',
since we see the same messages when we try to do a restore by hand.

Some notes:

1.	The TK50 *is* unit 5, and it *is* online.  This is the drive which
	was successfully addressed only minutes before, when the standalone
	system was booted from it.

2.	We've tried replacing this TK50 with another, identical, TK50 and 
	we see the same results.

3.	We've tried talking to the tape drive, using the 'mt' command 
	in the standalone system, and mt can't talk to the drive either.
	'mt status' results in:

	TK50 unit 5: receive diagnostics command failed.
	TK50 tape drive: residual = 0
	ds = 0
	er = 0

4.	Looking in /dev, we find that the proper device special files are 
	there.  Deleting these and re-creating them has no effect.

and finally, the coup de grace:

5.	We've booted an identical (working) system with the same tape, and
	found that it, too, can't talk to the local tape drive after booting.

	We're baffled, and the local DECpeople are baffled.  It's possible
that we may be able to lay hands on a later distribution tape eventually, but
for the moment we're stuck with this one.  (The system was originally installed
with an 'ULTRIX 32 v3.1D' tape, borrowed from another institution.)  Can 
anyone suggest a patch for the problem we're seeing?

			Any advice will be greatly appreciated.

			Bryan.
 
|||||||||||||||||||||||||||| Bryan K. Wright ||||||||||||||||||||||||||
|              University of Virginia Physics Department              |
|             (804) 924-6814 Office, (804) 293-7267 Home              |
| Internet: Bryan@gomez.phys.Virginia.EDU, Bitnet: bkw1a@Virginia.EDU |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| "If you take cranberries and stew them like applesauce they taste   |
|  much more like prunes than rhubarb does."  -- Groucho              |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||