> On Dec 20, 2017, at 10:47 AM, Dana Mitchell <[email protected]> wrote:
> 
> Mid 80's, we saw one at our DR Hotsite vendor Comdisco's Chicago location.  
> It was onsite for one of their customers,  it wasn't operational at the time 
> but we got to look inside.  They said it usually took them a few days before 
> a test to get it functioning properly.
> Dana
> 
> On Wed, 20 Dec 2017 14:21:38 +0000, PINION, RICHARD W. 
> <[email protected] <mailto:[email protected]>> wrote:
> 
>> Drifting slightly, ever see the IBM 3850 Mass Store device, or see it in 
>> action?

We managed to have both an MSS and a “drum” (separate and distinct devices).

The “drum” we had was in actuality memory, and it worked (details in next 
sentence).  I do not remember what year it was, but it had to be in the early 
1980’s. Our VP regularly was a person that bought things and thought later. The 
drum” that showed up in the computer room was a box about the size of two 
freezer units. We had not a clue it was coming it was just there. After the 
people unpacked the unit, there were two manuals one was an Engineer type 
manual (schematics), and one was a systems programmer type manual. I got to 
take it home that night and read the manual. 
Turned out that there box was just actually just memory, no real disk heads or 
anything that actually moved. It was in reality a 2305-2 look alike. You ran 
ICKDSF and clipped it and created a VTOC and after that it was pretty much like 
any disk drive. We ran a define page space and wallah it was a paging device. 
So we updated parmlib and at there next IPL we did a CLPA . it ran for a week 
without an issue. The *ONE* application that used it couldn’t tell the 
difference. The first 2 weeks went by and nothing happened. Then the summer 
from hell hit Chicago. 2-3 power outages a day. The first day it happened the 
system started up but could not access the device, so in order to get the 
system up and running we used another device as PLPA. AT this point the system 
was up and running so I was able to get the device online but it was empty. I 
had to clip the device and put a vote back on it and ran a define page space 
(at that time did not have the capability to add a page space). So in order to 
get it to be used by the system as a paging device we had to IPL again. This 
went on for several days and I was getting tir4d of going through the exercise. 
I finally complained to the VP and asked to talk with someone at the company 
that knew something about the device. I was talking with him and of course the 
power blipped and the entire system went down. I was muttering under my breath 
about the @#$%%^ device while I was talking and fixing the damn thing. He said 
he would call me back in a few minutes.  About 15 minutes later he calls me 
back and I explained what I was going through every time there was a power blip 
and he said he had never heard of this. I asked him to talk to the designer. 
About 1 hour later he calls me back and asks how it was going and I asked what 
did you find out? He told me that there was never a thought about when they 
designed the machine about power outages.  By that time I was really mad. About 
two hours later we had another power glitch. It was bad enough with the power 
glitches but to add on another IPL was pissing off the online people, we had 
over 1000 people that were taking calls from the public to order a magazine and 
they could not do so and their management (rightfully so, was getting mad). 
Three of us went up to the VP’s office and explained what was going on. He 
called up the salesman and asked for an explanation. The salesman put us on 
hold and made a call and came back on quickly and we were in conference mode. 
SO I explained yet again the issue with the damned device. The guy came on and 
actually said well that is in our next “release” a battery to keep the device 
from loosing its memory. Our VP told us to leave. The next day the order came 
down to schedule an IPL and delete the PLPLA and use another device). I did so 
and was happy to see the device leave. The device was boxed up and taken away. 
The following is interesting, IMO. After this fiasco we were told to have a 
meeting with the one user that needed the device, so we talked with him for 
about 30 minutes and one of us had the idea that the user should use VIO rather 
than the drum.
We ran a test, and it was a success. The user was happy, so we no longer needed 
the drum.

I managed to loose the 3850 story somehow and will try and tell it later.

Ed


----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to