Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > alt.folklore.computers > #227667

Re: The Fall Of OS/2

Show all headers | View raw

the IBM communication group was fiercely fighting off client/server and
distributed computing. Late 80s, a senior engineer in the disk division
got a talk scheduled at internal, communication group, world-wide
conference supposedly on 3174 performance ... but opened the talk with
statement that the communication group was going to be responsible for
the demise of the disk division. They were seeing a drop in disk sales
with data fleeing mainframe datacenters to more distributed computing
friendly platforms ... and had come up with a number of solutions.
However the communication group was constantly vetoing the disk division
solutions (with communication group corporate strategic
ownership/responsibility for everything that crossed datacenter walls).

note that workstation division had done their own cards for the PC/RT
(PCAT-bus) including 4mbit token-ring card. Then for the RS/6000
microchannel workstations, AWD was told they couldn't do their own
cards, but had to use PS2 microchannel cards. The communication group
had severely performance kneecaped the PS2 microchannel cards
... example was that the $800 PS2 microchannel 16mbit token-ring card
had lower card throughput than the PC/RT 4mbit token-ring card ... and
significantly lower throughput than the $69 10mbit Ethernet card.

trivia: The new IBM Almaden Research bldg had been extensively
provisioned with CAT wiring, presuming use for 16mbit token-ring ...
however they found that not only (CAT wiring) 10mbit Ethernet cards had
much higher throughput than 16mbit T/R cards, also 10mbit ethernet LANs
had higher aggregate throughput and lower latency than 16mbit T/R.

Also in the aggregate cost difference between the $69 Ethernet cards and
$800 16mbit T/R cards, Almaden could get nearly half dozen
high-performance tcp/ip routers ... each with 16 10mbit Ethernet
interfaces and ibm mainframe channel interfaces with options for T1&T3
telco interfaces, and various high-speed serial fiber interfaces.
Result was they could spread all the RS/6000 machines across the large
number of Ethenet (tcp/ip) lans ... with only a dozen or so machines
sharing a LAN.

Summer 1988, ACM SIGCOMM published study that 30 10mbit ethernet
stations ... all running low-level device driver loop constantly sending
minimum sized packets, aggregate effective LAN throughput dropped off
from 8.5mbit/sec to 8mbit/sec.

For fiber "SLA", RS/6000 had re-engineered & tweaked mainframe ESCON
... making it slightly faster (and incompatible with everything else)
... 220mbit/sec, full-duplex; so the only thing they could use it for
was with other RS/6000s. We con one of the high-speed tcp/ip router
vendors to add a "SLA" interface option to their routers .... giving
RS/6000-based servers a high-performance entre into distributed
computing envrionment.

In 1988, the IBM branch office had asked me if I could help LLNL
(national lab) standardize some serial stuff they were playing with,
which quickly becomes fibre-channel standards (FCS, including some stuff
I had done in 1980), initially 1gbit/sec, full-duplex, aggregate
200mbyte/sec. The RS/6000 SLA engineers were planning on improving SLA
to 800mbit/sec ... when we convince them to join the FCS standard
activity instead.

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

Back to alt.folklore.computers | Previous | Next — Previous in thread | Find similar

Thread

The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-06 00:31 +0000
  Re: The Fall Of OS/2 ted@loft.tnolan.com (Ted Nolan <tednolan>) - 2024-10-06 02:47 +0000
    Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-06 03:27 +0000
      Re: The Fall Of OS/2 Chris Ahlstrom <OFeem1987@teleworm.us> - 2024-10-06 06:55 -0400
        Re: The Fall Of OS/2 Peter Flass <peter_flass@yahoo.com> - 2024-10-06 15:58 -0700
          Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-06 23:24 +0000
          Re: The Fall Of OS/2 Bob Martin <bob.martin@excite.com> - 2024-10-07 04:16 +0000
        Re: The Fall Of OS/2 OrangeFish <OrangeFish@invalid.invalid> - 2024-10-07 10:47 -0400
          Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-07 20:37 +0000
            Re: The Fall Of OS/2 OrangeFish <OrangeFish@invalid.invalid> - 2024-10-08 13:08 -0400
              Re: The Fall Of OS/2 Bob Eager <news0009@eager.cx> - 2024-10-08 20:25 +0000
                Re: The Fall Of OS/2 Dave Yeo <dave.r.yeo@gmail.com> - 2024-10-08 19:45 -0700
    Re: The Fall Of OS/2 Dave Yeo <dave.r.yeo@gmail.com> - 2024-10-05 21:03 -0700
  Re: The Fall Of OS/2 Dave Yeo <dave.r.yeo@gmail.com> - 2024-10-05 20:57 -0700
    Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-06 13:40 +0000
      Re: The Fall Of OS/2 Dave Yeo <dave.r.yeo@gmail.com> - 2024-10-06 14:16 -0700
        Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-06 21:31 +0000
          Re: The Fall Of OS/2 John Ames <commodorejohn@gmail.com> - 2024-10-07 09:35 -0700
            Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-07 20:41 +0000
              Re: The Fall Of OS/2 John Ames <commodorejohn@gmail.com> - 2024-10-07 14:32 -0700
                Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-07 21:44 +0000
                Re: The Fall Of OS/2 John Ames <commodorejohn@gmail.com> - 2024-10-07 15:19 -0700
                Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-08 02:41 +0000
                Re: The Fall Of OS/2 John Ames <commodorejohn@gmail.com> - 2024-10-08 09:59 -0700
                Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-08 23:12 +0000
                Re: The Fall Of OS/2 John Ames <commodorejohn@gmail.com> - 2024-10-14 15:02 -0700
                Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-17 23:25 +0000
              Re: The Fall Of OS/2 Bozo User <anthk@disroot.org> - 2024-10-11 14:49 +0000
    Re: The Fall Of OS/2 Peter Flass <peter_flass@yahoo.com> - 2024-10-06 15:58 -0700
      Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-06 23:26 +0000
        Re: The Fall Of OS/2 Peter Flass <peter_flass@yahoo.com> - 2024-10-06 18:05 -0700
          Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-07 03:11 +0000
            Re: The Fall Of OS/2 John Ames <commodorejohn@gmail.com> - 2024-10-07 09:09 -0700
              Re: The Fall Of OS/2 Peter Flass <peter_flass@yahoo.com> - 2024-10-07 15:52 -0700
                Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-08 02:45 +0000
                Re: The Fall Of OS/2 Dave Yeo <dave.r.yeo@gmail.com> - 2024-10-08 19:56 -0700
            Re: The Fall Of OS/2 Peter Flass <peter_flass@yahoo.com> - 2024-10-07 15:52 -0700
              Re: The Fall Of OS/2 Dave Yeo <dave.r.yeo@gmail.com> - 2024-10-08 20:14 -0700
                Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-09 05:48 +0000
                Re: The Fall Of OS/2 Dave Yeo <dave.r.yeo@gmail.com> - 2024-10-10 19:28 -0700
                Re: The Fall Of OS/2 Lawrence D'Oliveiro <ldo@nz.invalid> - 2024-10-11 05:42 +0000
  Re: The Fall Of OS/2 jgd@cix.co.uk (John Dallman) - 2024-10-06 10:10 +0100
    Re: The Fall Of OS/2 Chris Ahlstrom <OFeem1987@teleworm.us> - 2024-10-06 07:01 -0400
    Re: The Fall Of OS/2 Lynn Wheeler <lynn@garlic.com> - 2024-10-06 09:08 -1000
      Re: The Fall Of OS/2 Lynn Wheeler <lynn@garlic.com> - 2024-10-06 13:33 -1000

csiph-web