Re: Comments to draft-ietf-ips-iSCSI-00.txt

["Jim Hafner/Almaden/IBM" <hafner@almaden.ibm.com>]

Matt,
Comments interspersed with <JIM> ...</JIM> and lots of your stuff clipped.

* Section 1.2.4, Login:
"A session is used to identify to a target all the connections with
a given initiator."

This is not true.  A session is used to identify
all the connections in a logical connection between the initiator
and the target.  There could be more than one session between an
initiator and a target, in which case a session does not identify
"all" the connections.

<JIM>
Actually, I think the statement is correct.  It depends on your definition
of "initiator".  In the SAM model (in particular, how SAM-2 is moving in
the multiport discussions) and how that fits with iSCSI, the initiator (as
far as the target is concerned) is defined (scoped?) by a single session.
It's the session that injects SCSI commands into the service delivery
subsystem and so it is the session which is the "application client" in the
SCSI Initiator Device (see t10/00-268r2 or maybe r3).  The target (in
principle) can't tell that the OSes (e.g.) driving two independent iSCSI
sessions might be the same.    This actually has implications for
reservations and such and may change (yet again) but that's the current
thinking and may be the safest approach.
</JIM>

-Matt Wakeley
Agilent Technologies

<JIM>
JIm Hafner, IBM
</JIM>