RE: iSCSI padding should be 0

["Eddy Quicksall" <eddy_quicksall@ivivity.com>]

Mike, you're not stepping on toes nor are you butting in. We need this kind
of input because no one on this reflector knows everything. Thanks for your
input.

In this case, I feel that MUST is appropriate and that if someone wants take
the time to check it then they are within their rights to do that. In fact,
your test tools SHOULD check it if we say MUST.

But, I don't think there would be a requirement for a target or initiator to
check it ... just to do it.

Eddy

-----Original Message-----
From: Fischer, Michael [mailto:Michael_Fischer@adaptec.com]
Sent: Friday, July 27, 2001 1:21 PM
To: 'eddy_quicksall@ivivity.com'; julian_satran@il.ibm.com
Cc: 'ips@ece.cmu.edu'
Subject: RE: iSCSI padding should be 0


Eddy,

This discussion is purely based on semantics and the definition of the word.
MUST indicates that it is REQUIRED.  Yes?  If it is REQUIRED that the
padding be set to 0 and it is NOT then that implementation is in violation
of the spec.  Yes?

You are correct that the RFC does not say that you must check it but it does
imply it (or at least that it is checkable), which is what I stated.  My
concern is that Julian suggested that the spec say MUST be 0 and that he did
NOT want people to check for 0.  To insure that it is not checked, the
statement MUST remain "SHOULD be 0."

In the production environment, the checking of whether or not the padding
has been set to zero may be bypassed for whatever reason, but in a testing
environment (guess where I work. :) ), I am sure that if someone read the
spec and implemented a test tool to verify that their implementation
conforms to the spec they could generate a bug report if the padding was not
0.

I am just trying to make sure that the spec is as unambiguous as possible
and I apologize if I am stepping on any toes.

Michael Fischer

-----Original Message-----
From: Eddy Quicksall [mailto:eddy_quicksall@ivivity.com]
Sent: Friday, July 27, 2001 9:32 AM
To: 'Fischer, Michael'; julian_satran@il.ibm.com
Subject: RE: iSCSI padding should be 0


The RFC says:
  MUST This word, or the terms "REQUIRED" or "SHALL", mean that the
  definition is an absolute requirement of the specification.

How does that say that I MUST check to be sure it has been done? Is that
requirement someplace else?


Eddy

-----Original Message-----
From: Fischer, Michael [mailto:Michael_Fischer@adaptec.com]
Sent: Friday, July 27, 2001 12:26 PM
To: 'Julian_Satran/Haifa/IBM%IBMIL'; eddy_quicksall@ivivity.com
Cc: ips
Subject: RE: iSCSI padding should be 0



Julian,

I hate to butt in with this minor point, I think you guys are doing a great
job in hammering out the specification and do not presume to know better.

However, according to RFC2119 the word MUST indicates that it is a
requirement of the specification and therefore implies that the receiver
will need to check the value.  If it MUST be zero and is not, then that
would indicate a protocol error.

Michael Fischer

-----Original Message-----
From: Julian_Satran/Haifa/IBM%IBMIL [mailto:julian_satran@il.ibm.com]
Sent: Friday, July 27, 2001 8:46 AM
To: eddy_quicksall@ivivity.com
Cc: ips
Subject: Re: iSCSI padding should be 0



Perhaps we should say MUST be sent as 0 and keep quiet about what the
receiver should  do (check for 0 - we don't want that).

Thanks,Julo

"Eddy Quicksall" <eddy_quicksall@ivivity.com> on 27-07-2001 18:18:33

Please respond to eddy_quicksall@ivivity.com

To:   Julian Satran/Haifa/IBM@IBMIL, ips@ece.cmu.edu
cc:
Subject:  iSCSI padding should be 0




Julian,

Section 2.1 says the padding should be 0. I guess that is correct because
one may not use CRC and therefore may not want to set them to 0. Wouldn't
it
be better if section 2.1 was more specific and mentioned when they must be
0
if there is a CRC. Also, I noticed at the UNH plug fest that at least one
person thought "should" meant "must". Therefore, I don't think it should
say
"should" ... I think it should not mention the 0'ness unless there is a CRC
present.

Also,

        transmission.  Padding bytes, when presents in a segment covered by
a
        CRC, have to be set to 0 and are included in the CRC.

should say "when present in".

Eddy_Quicksall@iVivity.com