On 11/22/2016 2:56 PM, Paul Berger wrote:
The channel adapter on the Series/1 had a rather large flaw, if you
did not disable the interface before shutting down the Series/1 it
would upset the channel it was attached to causing a flurry of channel
checks that could bring the host system to its knees. When I worked
in the IBM Toronto Lab we had two channel attached Series/1 machines
with 72MD diskette units that we used to create diskettes from images
sent to us and also to send diskette images. These Series/1s
pretended to be a 3270 control unit so that the MVS host system knew
how to talk to them.
the Ultimate system was the only Non IBM written supervisor / system
that ran on mainframes at the time. The supervisor or in Pick terms,
Monitor handled such things with code specific to Pick. I don't know
that it would have caused the system to stop, but I'd doubt it, as we
had a complete setup to test it with, and running with parts turned off
after system boot would have been tested.
All of the channels, 7171, Series 1, and Hyfas channels were all run in
a different mode than most anywhere in IBM, though I understand the mode
was supported w/o firmware mods on all systems. I'd have loved to have
heard the L2 and L3 calls if anyone called with complaints though. Our
support contract with IBM essentially made us L3 for everything we had,
but we had access to all PMR's as well as PMR's filed against systems
with our software. I say to all PMR's as the access has vastly
narrowed, but we had access for supporting our own VM/SP, so we could
search for problems and get APARs.
All of our systems were hooked up via a channel switch to three systems,
a 4381, 9121, and a 9370. All had some dedicated DASD, but unique items
such as a couple of tape drives, the 7171, and the Series one were all
on the switch to be moved to each system to see what happened.
We also conned IBM into a CHIM as well as set of hardware and software
to enable Microchannel machines to be either a channel, or to run
channel attached hardware.
Also, note that Mike Ross has a CHIM, which I am not sure he knows is a
CHIM.
http://www.corestore.org/360tester-1.jpg
At least the box we had matched this. We were hoping that IBM would
forget it when we started to get where our facility would shut down, as
we were the only account left in Orange County with 370 or 360 channels,
but the found it.
Also had a custom instruction trace firmware for the 4381 for being able
to trace some performances with deposits into lower memory.
thanks
Jim
