Re: iSCSI: multiple connection process

[julian_satran@il.ibm.com]

David,

Only for correctness - the SAME design team has discussed both the
symmetric and
asymmetric model and choose the symmetric because it did not have a good
enough
asymmetric solution;  that does not hold now anymore and that is why I
opened
this thread (should I say I am sorry?)

I have two objections to the direction you have taken:

- you considered objections that where voiced to you and not to the mailing
list;
   the community should judge if they are core or marginal; I would
appreciate if
   you could summarize them

- the key issues with any of the two approaches are understood - an
evaluation
   committee will only broaden the "discontent"

Julo



Black_David@emc.com on 25/09/2000 23:51:48

Please respond to Black_David@emc.com

To:   matt_wakeley@agilent.com, Black_David@emc.com, ips@ece.cmu.edu
cc:    (bcc: Julian Satran/Haifa/IBM)
Subject:  iSCSI: multiple connection process




> Exactly.  There has been NO consensus that multi-connection sessions
should be
> removed from the main specification (only your proposal to do that).
Just
> because we can't (yet) agree on something, does not mean "let's not do it
at
> all, or put it off until later."

Sorry, lack of rough consensus means one of those two.  Documents don't
leave the WG without rough consensus, so either the document gets delayed
for consensus, or the feature gets removed.  Needless to say, we don't have
rough consensus at the moment, nor do I see much in the way of common
ground emerging among the different points of view.

> We already had the offline design team and we came up with something that
> everyone is arguing about.  So let's have some more discussion on what it
is
> that people don't like, and what the requirements are, so that another
offline
> design team has something to work with.

Actually, offline design teams came up with both the Asymmetric and
Symmetric
models ... and now there's a third.  It appears to me that continued list
discussion is
an active version of "put it off until later" rather than a passive one.  I
could be
wrong, but someone will need to explain what will be different about
continued list
discussion that will lead to a consensus in a way that discussion since
Pittsburgh
has not.

The alternative process I have in mind is more of an evaluation team than a
design
team - form a team of people who publish evaluation criteria for
multiple-connection
sessions, evaluate drafts describing the various models against those
criteria,
and make a recommendation to the WG.  One possible way to make continued
progress on the list is to discuss what those evaluation criteria ought to
be
rather than asking an off-line team to write them.  Doing this really
requires someone
to track the criteria as they evolve - a volunteer for this task is hereby
solicited,
preferably someone who has not taken a position on this session issue.

Thanks,
--David

---------------------------------------------------
David L. Black, Senior Technologist
EMC Corporation, 42 South St., Hopkinton, MA  01748
+1 (508) 435-1000 x75140     FAX: +1 (508) 497-8500
black_david@emc.com       Mobile: +1 (978) 394-7754
---------------------------------------------------