RE: iSCSI: text commands that set mode parameters

["Robert Griswold" <rgriswold@Crossroads.com>]

Eddy:

I have found a couple of threads on this subject, and it seems that
quite a few others involved in T10 work agree with you on the fact that
mode page or setting manipulation outside of the mode commands is not
in-line with the scope of the iSCSI mapping.  I see that iSCSI needs to
be able to manipulate items such as Disconnect/Reconnect and BurstSize,
but if those settings live in Mode Pages, then iSCSI needs to use mode
commands to get or change them, right?

These threads discuss mode pages, but I am not sure I see any resolution
to this redundancy yet.  Can anyone tell me if there was a final
decision to modify the mode page language Eddy refers to?

http://ips.pdl.cs.cmu.edu/mail/msg04176.html
http://ips.pdl.cs.cmu.edu/mail/msg03519.html


Robert Griswold
Technologist
Crossroads Systems, Inc.
512-928-7272

 -----Original Message-----
From: 	Eddy Quicksall [mailto:eddy_quicksall@ivivity.com] 
Sent:	Thursday, June 07, 2001 4:06 PM
To:	ips@ece.cmu.edu
Subject:	iSCSI: text commands that set mode parameters

I'm not crazy about setting mode parameters from text commands because
the
mode pages are in logic which is outside of iSCSI.

 Section 3 - SCSI Mode Parameters for iSCSI says:
   The mode parameters can be set and retrieved by SCSI mode-set and
   mode-sense commands or with the iSCSI text commands and responses.
   The text commands offer the added convenience that at the end of the
   exchange the value selected is known to both parties.

I agree with the spirit of the last sentence above but I still think the
logic is misplaced.

Has anyone debated this yet?
Does anyone else share my opinion?
Is there a good reason to have the iSCSI layer do this?
Can we talk about this or is there a thread I should review?

Eddy