Re: iSCSI : Section 4. Login Phase.

[julian_satran@il.ibm.com]

answers in text - Julo

Santosh Rao <santoshr@cup.hp.com> on 13/03/2001 03:52:57

Please respond to Santosh Rao <santoshr@cup.hp.com>

To:   IPS Reflector <ips@ece.cmu.edu>
cc:
Subject:  iSCSI : Section 4. Login Phase.




Julian & All,

A couple of issues on Section 4 of the rev 05 iSCSI draft :

1) "The Login Final-Response can come only as a response to a Login
command with the F bit set to 1 or a Text Command with the F bit set to
1."

The above statement is not true and must be removed, since the
login-final response is used by targets when rejecting login. As per the
table in Section 2.11.4, the F bit is allowed to be set to 1 when the
status class is non-zero.


+++ will fix wording +++

2) "Operational parameters MAY be negotiated within or outside (after)
the login phase."

If text negotiation can be conducted outside of the login phase, what is
the effect of changes in text keys on the existing login ?
For ex, how does changes to keys like MaxConnections, UseR2T,
BidiUseR2T, ImmediateData, DataPDULength, FirstBurstSize, ITagLength,
etc affect their existing values negotiated during the current login ?

The side-effects of such changes and the scope of their establishment
must be documented. It may be simplest to prohibit such changes outside
of the login phase, for the sake of simplicity.

+++ most of the parameters you mentioned have the LO qualifier - i.e., they
can't be changed after the Login that establishes the session. Some are not
as I though that theuser may want them connection specific or even change
them (e.g., R2T) +++

Regards,
Santosh

 - santoshr.vcf