srp-r04.pdf

[Edward A. Gardner]

* From the T10 Reflector (t10 at t10.org), posted by:
* "Edward A. Gardner" <eag at ophidian.com>
*
Rob Elliott asked me to forward this to the reflector.  He hopes we might
reach quick agreement that target port identifiers should be 128 bits, the
same size as initiator port identifers.

-----Original Message-----
From: Edward A. Gardner <eag at ophidian.com>
To: Elliott, Robert <Robert.Elliott at COMPAQ.com>
Date: Thursday, May 10, 2001 2:31 PM
Subject: Re: srp-r04.pdf


>FCP-2 revision 7 is in public review, so that should be
>stable.

My concern is that it's an informative annex, not in a normative part of the
spec.  Traditionally that's been a symptom of last minute stuff that may
still be only partially baked.

All I'm really saying is that I want to be told those are the right words to
copy.  Or for the people who really care about bi-directional IO to review
them and tell me what needs to be changed.  I don't feel comfortable enough
with my understanding of the bi-directional stuff to make that judgement on
my own.  (I'm still part of the old school that thinks it doesn't belong in
SCSI. :-)

>The outcome of the March meeting was:
>* 128 bit initiator port ID (must be worldwide unique, formed
>from an EUI-64 plus 64 additional bits)
>* 64 bit target port ID (an EUI-64).
>
>I don't think it was accidental.  I requested 128 bits for
>both to make implementing targets in processor nodes easier,
>but the group (including you) didn't agree.  The rationale

We discussed the target port ID first and agreed to 64 bits (yes, I argued
for that).  Then we discussed the initiator port ID and you / Rob Haydt /
etc. prevailed in arguing for 128 bits.  If we'd discussed these in the
opposite order, I would have argued for the target to be 128 bits as well.
Yes, in principle they are independent, but my experience shouts that
keeping them symmetrical will turn out to be the right thing to do in the
long run.  That's why I say it was an accident, it happened because of the
way the agenda was laid out.  And by the time we'd finished, I was too worn
out to ask to go back and revisit it.

The change to the IU format is about 30 seconds, provided I'm already
generating a new version of the spec.  Literally, it takes longer to produce
the .pdf than to make that change to the IU format diagram.  And the
SRP_LOGIN_REQ diagram is the only place it appears in srp-r06 (it's much
bigger impact on your annex).

Edward A. Gardner               eag at ophidian.com
Ophidian Designs                719 593-8866 voice
1262 Hofstead Terrace           719 593-8989 fax
Colorado Springs, CO  80907     719 210-7200 cell


*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org