Re: Filesystems

[Anne & Lynn Wheeler <lynn@garlic.com>]

jeffj@panix.com (Jeff Jonas) writes:
> I consulted to IBM for an AIX project.
> IBM fostered a horrible "not invented here" mentality.
> IBM had hired Convergent to make 2 AIX ports:
> for the PS/2 and the 3090 mainframe.
> They were tightly coupled into clusters and the TCF (totally confusing facility).
> No matter how much I wanted to dislike AIX,
> I saw many merits and unique features.
> I suspect that IBM killed that project (even though they owned it and PAID for it)
> since they wanted to re-invent it all with "native" code.
> Surprisingly, AIX for the Risc-6000 succeeded since it was written
> by the Austin Texas facility and was apparently exempt from the NY politics.

there were several ... totally different unix projects.

there was the tss/370 which modified the tss kernel so that at&t unix
layer sat on top .... I don't know if that ever saw deployment other
than inside of at&t. There was also a lot of effort to do a similar
at&t port on top of VM ... but I don't know if that ever saw any
customer installations.

there was the (at&t) pc/ix port by interactive for the pc. when the
opd romp displaywriter project was killed ... the austin group quickly
retargeted for unix workstation and got interactive to do a port.
however the claimed that it would be faster if interactive ported to
an abstract machine layer (rather than interactive having to learn the
details of the romp hardware) ... and so a lot of the pl.8
displaywriter people were put to work writing the vrm (in pl.8).
This was shipped (combined vrm & interactive) as aix for pc/rt
(reworked displaywriter).

palo alto had been working on bsd port for the 370. with the advent of
the pc/rt ... they retargeted their work to the pc/rt ... which
eventually resulted in something called AOS (bsd for the pc/rt). They
also somewhat disproved the original austin assertion having done a
bsd port to the native pc/rt in possibly 1/10th the time/effort it
took interactive to port to the VRM abstraction (and possibly 1/100th
the time/effort for the combined interactive port plus VRM
development).

AIX for the RS/6000 had AIX for the PC/RT reworked .... eliminating
the pl.8 code and calling it AIX V3.

Palo Alto ... dating back to the early '80s had been doing work with
both Berkeley and BSD as well as UCLA and Locus. In the early '80s
Palo Alto had Locus running on S/1 and some 68000 machines in
distributed Locus environment. This was ported to 370 and PS/2s and
was the officially shipped AIX/370 and AIX/ps2. This palo alto/locus
was totally different "aix" than the austin "aix".

Austin for AIX V3 had developed the journal file system where the
filesystem was slightly re-organized so that all of the metadata was
in area of virtual memory that was sort of designated "database
memory". This was sort of from one of the early 801 design activities
where they wanted to show hardware assit for transactional database
systems. Basically hardware assist to track transaction lines (about
the size of cache lines) got modified ... and then at commit ... the
commit code could run around and find all the modified lines/data and
log it. Then you start the writes ... and if there was a failure
... restart could use the log to consistently roll-forward the
database.

Early in the rs/6000/aix time-frame, Palo Alto had a project to port
AIXv3 to different platforms. One of the inhibitors was the JFS
dependency on the 801 database memory hardware. Palo Alto undertook to
go thru the unix filesystem code and insert "log" calls whenever
metadata was being modified (these "log" calls weren't necessary in
801 since all instances of data modification was magically identified
by the hardware). It did have the downside with their relationship
with austin (which was already strained .... in part because of
showing them up with the bsd/aos port on the pc/rt ... as well as the
locus activity shipping as aix/370/ps2) ... since the "log call"
version was measurably faster than the original version using the 801
hardware. turns out in the hardware version, commit had to scan the
lines in the database region for modified lines ... if the total size
of metadata was significantly larger than the nominal lines modified
per commit, the commit scanning took much longer than the overhead of
the direct log calls.

For OSF there was an attempt at merge of many of these different
threads ... in part for DCE ... bringing together mit, cmu, berkeley,
ucla with various apollo distributed support, cms andrew distributed,
the locus distributed stuff, austin aix distributed filesystem, etc.

aixv2    .... iteractive at&t on top of vrm for pc/rt
aos      .... bsd for pc/rt
aixv3    .... aixv2 with vrm eliminated for rs/6000
aix/370  .... ucla locus
aix/ps2  .... ucla locus

there were misc. other bsd & at&t ports along the way for vm/370 and
tss/370.

misc. past "aix" posts
http://www.garlic.com/~lynn/96.html#4a John Hartmann's Birthday Party
http://www.garlic.com/~lynn/99.html#2 IBM S/360
http://www.garlic.com/~lynn/99.html#36 why is there an "@" key?
http://www.garlic.com/~lynn/99.html#63 System/1 ?
http://www.garlic.com/~lynn/99.html#64 Old naked woman ASCII art
http://www.garlic.com/~lynn/99.html#65 Old naked woman ASCII art
http://www.garlic.com/~lynn/99.html#66 System/1 ?
http://www.garlic.com/~lynn/99.html#129 High Performance PowerPC
http://www.garlic.com/~lynn/2000.html#49 IBM RT PC (was Re: What does AT stand for ?)
http://www.garlic.com/~lynn/2000.html#64 distributed locking patents
http://www.garlic.com/~lynn/2000c.html#8 IBM Linux
http://www.garlic.com/~lynn/2000d.html#65 "all-out" vs less aggressive designs (was: Re: 36 to 32 bit transition)
http://www.garlic.com/~lynn/2000e.html#27 OCF, PC/SC and GOP
http://www.garlic.com/~lynn/2001.html#44 Options for Delivering Mainframe Reports to Outside Organizat ions
http://www.garlic.com/~lynn/2001.html#49 Options for Delivering Mainframe Reports to Outside Organizat ions
http://www.garlic.com/~lynn/2001f.html#1 Anybody remember the wonderful PC/IX operating system?
http://www.garlic.com/~lynn/2001f.html#20 VM-CMS emulator
http://www.garlic.com/~lynn/2001f.html#22 Early AIX including AIX/370
http://www.garlic.com/~lynn/2001i.html#21 3745 and SNI
http://www.garlic.com/~lynn/2001i.html#30 IBM OS Timeline?
http://www.garlic.com/~lynn/2001j.html#20 OT - Internet Explorer V6.0
http://www.garlic.com/~lynn/2001l.html#5 mainframe question
http://www.garlic.com/~lynn/2001l.html#8 mainframe question
http://www.garlic.com/~lynn/2001l.html#17 mainframe question
http://www.garlic.com/~lynn/2001l.html#19 mainframe question
http://www.garlic.com/~lynn/2001l.html#50 What makes a mainframe?
http://www.garlic.com/~lynn/2001n.html#23 Alpha vs. Itanic:  facts vs. FUD
http://www.garlic.com/~lynn/2002b.html#29 windows XP and HAL: The CP/M way still works in 2002
http://www.garlic.com/~lynn/2002b.html#36 windows XP and HAL: The CP/M way still works in 2002
http://www.garlic.com/~lynn/2002d.html#31 2 questions: diag 68 and calling convention
http://www.garlic.com/~lynn/2002g.html#2 Computers in Science Fiction
http://www.garlic.com/~lynn/2002g.html#39 "Soul of a New Machine" Computer?
http://www.garlic.com/~lynn/2002h.html#19 PowerPC Mainframe?
http://www.garlic.com/~lynn/2002h.html#65 Bettman Archive in Trouble
http://www.garlic.com/~lynn/2002h.html#79 Al Gore and the Internet
http://www.garlic.com/~lynn/2002i.html#54 Unisys A11 worth keeping?
http://www.garlic.com/~lynn/2002i.html#81 McKinley Cometh
http://www.garlic.com/~lynn/2002j.html#36 Difference between Unix and Linux?
http://www.garlic.com/~lynn/2002m.html#21 Original K & R C Compilers
http://www.garlic.com/~lynn/2002n.html#67 Mainframe Spreadsheets - 1980's History
http://www.garlic.com/~lynn/2002o.html#11 Home mainframes
http://www.garlic.com/~lynn/2002o.html#40 I found the Olsen Quote
http://www.garlic.com/~lynn/2002p.html#45 Linux paging
http://www.garlic.com/~lynn/2003c.html#49 Filesystems
http://www.garlic.com/~lynn/2003d.html#8 IBM says AMD dead in 5yrs ... -- Microsoft Monopoly vs. IBM


and then different lineage was gold (aka A-U) shipped as UTS:
http://www.garlic.com/~lynn/99.html#190 Merced Processor Support at it again
http://www.garlic.com/~lynn/99.html#191 Merced Processor Support at it again
http://www.garlic.com/~lynn/2000c.html#8 IBM Linux
http://www.garlic.com/~lynn/2000f.html#68 TSS ancient history, was X86 ultimate CISC? designs)
http://www.garlic.com/~lynn/2000f.html#69 TSS ancient history, was X86 ultimate CISC? designs)
http://www.garlic.com/~lynn/2000f.html#70 TSS ancient history, was X86 ultimate CISC? designs)
http://www.garlic.com/~lynn/2001e.html#19 SIMTICS
http://www.garlic.com/~lynn/2001l.html#18 mainframe question
http://www.garlic.com/~lynn/2002e.html#46 What goes into a 3090?
http://www.garlic.com/~lynn/2002i.html#63 Hercules and System/390 - do we need it?
http://www.garlic.com/~lynn/2002n.html#54 SHARE MVT Project anniversary
http://www.garlic.com/~lynn/2002n.html#58 IBM S/370-168, 195, and 3033
http://www.garlic.com/~lynn/99.html#2 IBM S/360

-- 
Anne & Lynn Wheeler | http://www.garlic.com/~lynn/ 
Internet trivia 20th anv http://www.garlic.com/~lynn/rfcietff.htm