Pat
It could well be that we are both right in that SNA architecture has
highjacked a set of initials used by what can be characterised as that part
of pre-SNA NCP which interworked with TCAM and VTAM basic mode and which
became called in later NCPs as the BSC-start-stop processor - or something
like that - and which was the part which interfaced with NTO. That use of
BTU I had completely forgotten about if I was ever acquainted with it.
In terms of early NCP documentation there was a manual in the pre-program
product NCP 5 days which gave very handy descriptions of what actually went
on inside NCP. That would have been the one to hang onto if you needed good
information on the BSC-start-stop processor part of NCP. A fairly close
second however was the Don Buckingham's BSC/start-stop NCP redbook.
The BTU which I thought I recalled and which was of so little importance
that I skipped it in my SNA Concepts presentation is *not* what I suggested
in the previous post - so you are quite right to have me revisit the topic!
Googling a bit more, I found the following, an explanation a bit closer to
home than Microsoft and Cisco and a bit closer to your heartland as well!
From NetView(R) Performance Monitor Reference, SH19-6965-06, Glossary:
<quote>
basic transmission unit (BTU)
In SNA, the unit of data and control information passed between path control
components. A BTU can consist of one or more path information units (PIUs).
See also blocking of PIUs.
...
blocking of PIUs
In SNA, an optional function of path control that combines multiple path
information units (PIUs) in a single basic transmission unit (BTU).
Note: When blocking is not done, a BTU consists of one PIU.
</quote>
What this means is that BTU is an alternative, the correct alternative, to
what I always called simply a PIU for subarea traffic and a PIU segment for
peripheral node traffic - and subarea traffic when segmenting support
appeared there also relatively recently. When talking about blocking, I
guess I should have been strict and described the string of PIUs as a BTU.
Thus the use of BTU in the SNA Formats XID fields is correct. One famous
field - because it means that specifying MAXDATA on a PU statement is a
waste of time when format 3 XIDs are used - is in bytes 21 and 22 "Maximum
BTU length that the XID sender can receive". Thus the BTU corresponds to
what I used to call the PIU segment.
Since I have a presentation where I describe NCP mechanisms in as much
detail as I could track down, including peripheral node segmentation, I
found that I had managed to describe BTU correctly there in conjunction with
peripheral node segmenting - almost certainly having copied the essence of
the description from Don Buckingham's SNA/NCP redbook (with Wolfgang
Singer). I always did rely on my foils to see me through a presentation!
Later I inserted a diagram and notes to cover the introduction of "blocking"
but somehow managed not to use the term BTU.
Incidentally whenever you mention FIDx, you are talking about a type of PIU.
Finally, Googling on, I came across what may be a doctoral thesis - if I
took the trouble to exercise my German somewhat - which includes a
description of SNA - in English - and which has a handy diagram explaining
all the SNA data units - including BTU of course. It missed the segmentation
aspect but even a German doctoral thesis can't be perfect!
Chris Mason
P.S. I do have a bunch of little old handbooks of various vintages in the
basement somewhere. I'll have to look them over for "basic mode" formats and
manipulations.
----- Original Message -----
From: "Patrick O'Keefe" <[EMAIL PROTECTED]>
Newsgroups: bit.listserv.ibm-main
To: <[email protected]>
Sent: Thursday, September 20, 2007 9:22 PM
Subject: Re: Humor a little OT
On Thu, 20 Sep 2007 03:06:33 +0200, Chris Mason
<[EMAIL PROTECTED]> wrote:
...
[3] From the Cisco site: "BTU - (basic transmission unit) In SNA, the
unit
of data and control information passed between path control
components. A
BTU can consist of one or more path information units." SNA Formats
mentions
BTU from time to time, in XID fields for example, but nowhere spells
out
what it means!
...
I don't have any of my old NCP doc and am relying on VERY old
memories, but if I remember correctly, those descriptions, while
not exactly wrong, are pretty wide of the mark. The BTU support
was from the pre-ACF days, and still lived in those old parts of the
NCP that supported NTO, etc. It had the same data as a FID0 but
in a different order. Inbound data transfer went BTU->FID0->FID4.
(Or FID1 before VTAM and NCP (v1) r3.) If you still have an old NCP
handbook I think you'll find the BTU in there.
Jees. I can't believe I'm brave/stupid enough to tell Chris Mason
anything about NCP. Keep in mind that I'm trying to remember
what I learned in an NTO class 25 years ago. And I have a very
faulty memory. :-)
Pat O'Keefe
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
