Text preview for : B233 MVS For Small Systems Users; Lauzon.pdf part of IBM B233 MVS For Small Systems Users; Lauzon IBM share SHARE_61_Proceedings_Volume_1_Summer_1983 B233 MVS For Small Systems Users; Lauzon.pdf
Back to : B233 MVS For Small System | Home
.:---
SESSION REPORT i@SHAREt;
-:......-=-
61 B233 MVS For. Small SYstems Users /GO The 3031-5 operates with MYT under VM/SP. The 3031-8
SHARE NO. SESSION NO. SESSION TITLE ATTENDANce
runs with MVS/SP 1.3 in native mode. We are in our last stages of
MVS Proiect John Cotto CHP an MVT to MVS conversion which should be completed by the end of
SESSION CHAIRMAN INST.CODE
the year. At that time, both CPU's will run MVS/SP 1.3, with
PROJECT
plans for the larger machine to run under VM/SP.
55 Merritt Blvd. Trumbull, CT 06611 (203) 377-2300
SESSION CHAIR~N'S COMPANY. ADDRESS. and Ptl0NE NUMBER Come September, the MVS machine will support 15 CICS
terminals for a Registrar's system, 15 TSO/SPF terminals for staff
program development, some 60 to 80 concurrent WYLBUR (ASCII)
MVS FOR SMALL SYSTEMS USERS terminals for our academic community, as well as a half dozen
batch initiators. Currently, the WYLBUR terminals and half the
batch initiators are not being supported under MVS. They are
Roger C. Lauzon presently on the 3031-5 under MVT along with 30 data base termi-
nals, more WYLBUR terminals, and other batch initiators.
Computer Centre
University of Windsor The 3031-8 was installed this summer. Before that time,
Windsor, Ontario the 3031-5 ran VM/SP with MVT as the favoured production virtual
Canada N9B 3P4 machine, and MVS/SP 1.3 as a 'quasi-production' system. This was
(519) 253-4232, X 648 done only to expedite the conversion to MVS, and it is not a
recommended practice.
Installation Code: UOW
MVS USER EXPERIENCE
MVS project Currently, the 3031-8 supporting CICS, TSO/SPF and
batch initiators runs very well. The terminal response time is
.... Session B233
August 25, 1983
excellent. Paging is at a minimum. But when we tried the same
functions on the 3031-5, the batch initiators affected the terminal
,....
00
response times quite dramatically. But we feel we could have
improved the situation somewhat if the disks were not so limited
INTRODUCTION and would have been set up just for that MVS system, not for VM
and MVT as well. The difference of 3 megabytes of real memory
The University of Windsor is a small university with was significantly apparent to us as well. Come September, when
some 8,500 full-time and 4,500 part-time students. Almost all we add all the concurrent WYLBUR terminals to MVS on the 3031-8
academic and administrative computing is performed in one data system, we feel we will have to back off most of the batch
processing centre run by the Computer Centre department that is initiators during the day shift to achieve good on-line response.
open seven days a week, twenty-four hours a day. As well, our on-line usage is not only from 9AM - 5PM, since the
students use the WYLBUR terminals around the clock. To achieve
The Computer Centre is a small to intermediate size MVS good on-line response, some other users, normally batch, must be
user. This paper will describe the hardware and software environ- cut back.
ment in the Computer Centre, our experience with MVS, and our con-
cerns as a small MVS user. When we ran both MVT and MVS under VM on the 3031-5,
the on-line response on MVS was terrible. Two major factors
affected this: we favoured MVT as our production machine, and
ENVIRONMENT we could not run MVS in a V=R environment since we only had 5
megabytes of real memory. Once we tried running MVS in a 4
We currently have two IBM 3031's, one with 5 megabytes megabyte V=R area, which allowed fairly good response on the
and the other one with 8 megabytes of memory. They eaph have MVS side, but MVT died.
three strings of DASD, all 3350 technology except one string of
3330-11's which can be sharable but normally is not. Disk stor-
age is limited; for one CPU we have 4,400 megabytes, whereas the CONCERNS AS A SMALL USER
other contains 4,800 megabytes. Both CPU's share