check the iocp manual - it's  the CSS-ID (the logical channel subsystem id).

http://www-01.ibm.com/support/docview.wss?uid=isg2de3388ad7e19bffc85257768003f170e&aid=1

page 68 and of course page "303", which says the 2817 has max of 3 for css-id, 
and a max of 2 for subchanel set id.

p.s. i loved the history lessons, most of them being fairy tales. and dark 
distant foggy  memories.
The truth though is more fascinating .

it's sad people can't let go.... device addressing was 3 digits(hex) .  let it 
go...
it's all now chanell-sets and iodf device numbers. the old days are gone, and 
should be forgotten.
let it go, ok?

A few yerars back IBM was well aware of the "device number running out"problem, 
they addressed in a few ways,
1) more "data" per the "addressed-device" - notably the dasd devices.
2) add channel sets and 
3) add virtual "layering".
 The thought was that this would allow the HCD/IOCP/HSA areas to be managed 
better(controlled).
and circumvent the  "software intrusion into created 3or4-byte addressing" i.e. 
5 hex digites need 3 bytes to live in.
and 3 bytes would upset the world of ZOS, (JCL,control blocks, etc etc).
Imagine the 2 byte fields "abcd" (that reresented 4 digits in 2 bytes) 
expanding to larger size as well.
it was a nightmare, to change pgms that dated back years and decades to 
"evaluate and change", let alone user-pgms
sysprogs wrote.  a lot of money here.  best solution- keep 4 char device 
numbers asis, add a new plane of existance
and let the hotshot sysprogs deal with it. first confusion, then awareness 
(i.e. read newera software share presentation handout --session 10471 last 
march titled "IODF, etc.",
  and/or other share sessions in last 20th century hosted by IBMers in a 
freeforall on this topic, it's been over a decade
-- where were you?)




so, it was decided to go another route instead -   the one they implemented 
starts at the IOCP level, and will
one day be "better known".

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

Reply via email to