[email protected] (Anne & Lynn Wheeler) writes:
> lots of customers had been convinced to order 360/67 to run the
> "official" ibm virtual memory operating system, tss/360 ... however
> tss/360 had difficulty making it to production level ... so many
> locations just ran the machine as 360/65 with os/360. 
> http://en.wikipedia.org/wiki/TSS/360

re:
http://www.garlic.com/~lynn/2014d.html#16 [OT ] Mainframe memories
http://www.garlic.com/~lynn/2014d.html#22 [OT ] Mainframe memories
http://www.garlic.com/~lynn/2014d.html#23 [OT ] Mainframe memories

as undergraduate in the 60s, I got to rewrite a lot of cp67 (after it
being installed in at the univ in jan1968). I had to compete with the
IBM SE working on tss/360 for weekend dedicated stand-alone time (the
rest of the time, the machine ran os/360).

one of the things I did was dynamic adaptive resource manager
... periodically called "fair share" scheduler because a default
resource management policy was "fair share" ... some past posts
http://www.garlic.com/~lynn/subtopic.html#fairshare

... and would ridicule the table driven scheduler in tss/360.

the IBM SE and I did a synthetic workload benchmark ... simulating user
think time doing fortran program edit, compile and execute ...  he ran
it on tss/360 with four simulated users ... I ran it on cp67/cms (on
identical hardware) with 35 simulated users ... getting better response
and throughput (than he did with four simulated users).

one of tss/360 features was "single level store" ... basically
filesystem was treated as extension of memory access ... which had
very poor throughput. For whatever reason, the tss/360 "single level
store" design was carried over into the future system effort
http://www.garlic.com/~lynn/submain.html#futuresys

after joining the science center, I did a form of "single level store"
for CMS as paged-mapped filesystem ... but fixing most of the throughput
problems I had observed in tss/360 ... and was part of the reason that I
would periodic ridicule what they were doing in FS (considering both my
resource management and my paged mapped filesystem significantly better
than what they were doing).

this was included in stuff that I later migrated from cp67 to vm370
... but the page mapped filesystem was not one of the things included
for release to customers .... possibly because of the bad rep that
"single level store" got in both tss/360 and future system. misc. past
posts
http://www.garlic.com/~lynn/submain.html#mmap

one of the things that I did have lots of problems with was supporting
position independent code (mentioned in the tss/360 wiki article) ...
constantly having to hack code to make in position independent
http://www.garlic.com/~lynn/submain.html#adcon

-- 
virtualization experience starting Jan1968, online at home since Mar1970

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

Reply via email to