[hacknews] disk follies, continued once more

henry@utzoo.UUCP (Henry Spencer) (11/09/83)

The intent today was to copy the system to the loaner pack on drive A,
switch drive numbers, and start running from A.  (All this is intended
to let us take drive B down for maintenance, and to let Terabit take
pack UNIXA4 away for refurbishing.)  Well, it didn't quite work out
as intended.

The pack-to-pack copy went quite smoothly, taking only about 15 minutes.
No problems at all.  Then I powered the system down, switched radial
cables, switched identity plugs, and tried to boot from drive A.  It
went ok until the point where the system first tries to write on its
pack, just after "mem =".  At this point the drive lit up SELECT LOCK
and refused to cooperate.  The disk diagnostic gave the same problem
(after a bit of floundering because I specified the wrong drive -- no
harm done, since I'd stopped drive B as a routine precaution).  I then
powered down, switched everything back, and tried the diagnostic again.
Same complaint (still on drive A, with B stopped).  I booted Unix up
from B, no problems.  Paused briefly to try the pack-to-pack copy again;
now it too gave SELECT LOCK on drive A.  Abandoned the effort and brought
the system up on B.

This smells like a cabling problem, since everything worked fine until
I switched cables.  Maybe we need a new radial cable for drive A.  Argh.

(Minor note on procedures:  detecting SELECT LOCK can get a bit complex
since most of the bulbs in our disk-front lights are burned out.  We in
fact have only one good bulb out of the six.  I've been moving it around
so that it is in the SELECT LOCK position on the drive of interest.)
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry