Entire RAID set non-reporting

I was gone from my office for 9 days doing a CISC to RISC migration which went
beautifully -- jobs that used to take hours now literally take minutes.  It's
great.  Plus now I get to code to V4R5 features instead of V3R2, and get to take
advantage of all that great stuff Hans and Company have been adding to RPG.

But returning to my home office, my personal 40S does not want to IPL normally.
I typically leave it running 24/7, but turned it off since I was going to be
gone so long.  It was last shut off in October, when I was gone for over 2
weeks.

My 40S-2109 has ten 6606 drives -- two mirrored on the MFIOP for the load
source, and eight in a RAID set attached to a 6502 controller in a 7117
expansion chassis.  All are in ASP 1.  It is running V4R3, although I had been
planning to install V4R5 today.  Guess that will have to wait for now...

An IPL set to B N results in the attention light with the following SRC's:

 11-3  A600 5090
 12-3  09B0 0160
 13-3  0020 0000
 14-3  0000 0150
 15-3  0300 FFFF
 16-3  6606 5074
 17-3  0000 A0C6
 18-3  0008 0000

My V4 Service Functions manual (SY44-5902) gives me most, but not all, of the
explanation of these codes.  In particular, it does not tell me what the 5090
unit reference code means in the A600 xxxx LIC interruption SRC.  Nor does it
tell me what "cause code" 0020 means in function 13-3.

Function 14-3 flags 0150 point to my 6502 card and the eight 6606 drives on it.
I am less sure of function 15-3's unit address of 0300.  If that is the drive
DST reports at I/O bus 0, controller 3, device 0 then it would be storage unit
location 12 and the third of the eight drives non-reporting.  The serial number
A0C6 in function 17-3 doesn't seem to match anything or at least not in the 6502
controller and the 6606 drives.

A manual IPL stops at IPL step C600 4008 (during sys config testing) and puts up
a display on the console reporting some disks as missing.  Displaying the detail
shows all 8 drives in the expansion unit as missing.

I wouldn't suspect 8 drives to exhibit the infamous 6606 "stiction" problem at
the exact same time, and not the other two attached to the MFIOP.  So my
inclination is to suspect the 6502 controller or the expansion feature power
supply.  But I'm measuring 12.24 and 5.04 vots on the power leads, and DST shows
the status of the 6502 as "Operational".

With the covers off so I can listen closer, I can hear when the first two drives
on the MFIOP are powered up sequentially.  But then I don't even hear it trying
to start the other 8 drives in turn.  Also, tapping lightly on the drives during
this time when I'd expect them to spin up seems to not have on effect either.

I don't have a spare 65x2 controller laying around to try swapping it out even
though it reports as operational under DST.  How do I narrow this down more?

TIA,
Doug
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---