Re: [storm] Drafts of new iSCSI specs

[<[e-mail]>]

Follow-up on this - I believe that we should stick with the two MAY-use requirements that were in RFC 3720.

Thanks,
--David


> -----Original Message-----
> From: [e-mail] [mailto:[e-mail]] On Behalf Of [e-mail]
> Sent: Wednesday, August 24, 2011 7:10 PM
> To: Banta, Andy (VMWare); [e-mail]
> Cc: Banta, Andy (VMWare); Huang, Kun (VMWare); Tomar, Nagendra (VMWare); Sreekanti, Kumar (VMWare);
> Aithal, Prasanna (VMWare); Goyal, Neeraj (VMWare)
> Subject: Re: [storm] Drafts of new iSCSI specs
> 
> Andy,
> 
> Many thanks for looking at the consolidated draft.
> 
> > In section 9.2, authentication is now required (MUST rather than MAY).
> > I'm quite sure this isn't going to fly with everyone.  CHAP is rarely
> > used in production environments, and until there's some distributed
> > key authentication method, I don't think many customers are going
> > to be interested.
> 
> Indeed it does, good catch, thank you.
> 
> The offending text in the consolidated draft is:
> 
> 9.2. In-band Initiator-Target Authentication
> 
>   During login, the target MUST authenticate the initiator and the
>   initiator MAY authenticate the target.
> 
> RFC 3720 had this text instead:
> 
> 8.2.  In-band Initiator-Target Authentication
> 
>    During login, the target MAY authenticate the initiator and the
>    initiator MAY authenticate the target.
> 
> Thanks,
> --David
> ----------------------------------------------------
> David L. Black, Distinguished Engineer
> EMC Corporation, 176 South St., Hopkinton, MA  01748
> [phone]             FAX: [phone]
> [e-mail]        Mobile: [phone]
> ----------------------------------------------------
> 
> > -----Original Message-----
> > From: Andy Banta [mailto:[e-mail]]
> > Sent: Monday, August 15, 2011 1:02 PM
> > To: Black, David; [e-mail]
> > Cc: Banta, Andy (VMWare); Sreekanti, Kumar (VMWare); Huang, Kun (VMWare); Aithal, Prasanna (VMWare);
> > Tomar, Nagendra (VMWare)
> > Subject: Re: Drafts of new iSCSI specs
> >
> > On Thu, Aug 04, 2011 at 12:08:38AM -0400, Black, David wrote:
> > > Andy,
> > >
> > > Can I interest you or anyone else at VMware in taking a look at these
> > > specs - it
> > > shouldn't involve a lot of time.  I can extend the deadline for input
> > > past VMworld if that helps.
> >
> > David, Folks,
> >
> > I'm the vSphere iSCSI development tech lead at VMware.
> >
> > I got a chance to look at the first one.  A few comments:
> >
> > In section 9.2, authentication is now required (MUST rather than MAY).
> > I'm quite sure this isn't going to fly with everyone.  CHAP is rarely
> > used in production environments, and until there's some distributed
> > key authentication method, I don't think many customers are going
> > to be interested.
> >
> > If this RFC gets approved as written, it will regularly be violated at
> > this clause.
> >
> > We are interested in coming up with a pluggable authentication method,
> > but I don't think that will change the spec in any way.
> >
> > I don't have any other specific comments based on the changes, but
> > have not gone through the spec in detail.  I'll take a look at the
> > second spec in the next few days.
> >
> > Thanks,
> > Andy
> > [e-mail]
> >
> > > Thanks,
> > > --David
> > >
> > > > -----Original Message-----
> > > > From: Black, David
> > > > Sent: Friday, July 15, 2011 2:01 AM
> > > > To: Banta, Andy (VMWare)
> > > > Cc: Black, David
> > > > Subject: FW: Drafts of new iSCSI specs
> > > > Importance: High
> > > >
> > > > Hi Andy,
> > > >
> > > > It was good to see you at EMC World, even if only briefly.
> > > >
> > > > I'm one of the co-chairs of the IETF storm (STORage Maintenance), where
> > > > new drafts of the iSCSI specifications are close to completion (they're
> > > > in Working Group Last Call).  We're looking for review and feedback from
> > > > iSCSI implementers, and I was hoping that you could take a look at this
> > > > on behalf of ESX's iSCSI implementation.
> > > >
> > > > There are two new iSCSI drafts:
> > > >
> > > > (1) iSCSI Protocol (Consolidated), draft-ietf-storm-iscsi-cons-03
> > > > 	http://datatracker.ietf.org/doc/draft-ietf-storm-iscsi-cons/
> > > >
> > > > This draft consolidates several existing iSCSI RFCs, primarily RFC 3720
> > > > and RFC 5048, and removes some unimplemented features - the result should be
> > > > backwards-compatible with existing implementations.  As the draft is several
> > > > hundred pages in length, I don't expect you to read it in its entirety, but
> > > > could you look at this summary of what's been changed and see whether it's
> > > > reasonable?:
> > > >
> > > > 2.3. Summary of Changes
> > > >
> > > >    1)     Consolidated RFCs 3720, 3980, 4850 and 5048, and made the
> > > >            necessary editorial changes
> > > >    2)     iSCSIProtocolLevel is specified as "1" in section 13.24, and
> > > >            added a related normative reference to [iSCSI-SAM] draft
> > > >    3)     Markers and related keys were removed
> > > >    4)     SPKM authentication and related keys were removed
> > > >    5)     Added a new section 13.25 on responding to obsoleted keys
> > > >    6)     Have explicitly allowed initiator+target implementations
> > > >            throughout the text
> > > >    7)   Clarified in section 4.2.7 that implementations SHOULD NOT
> > > >          rely on SLP-based discovery
> > > >    8)   Added UML diagrams, and related conventions in section 3
> > > >    9)   FastAbort implementation is made a "SHOULD" requirement in
> > > >          section 4.2.3.4 from the previous "MUST" requirement.
> > > >    10) Clarified in section 6.2 that validity of NotUnderstood
> > > >          response depends on iSCSIProtocolLevel
> > > >
> > > > (2) iSCSI SAM features, draft-ietf-storm-iscsi-sam-03
> > > > 	http://datatracker.ietf.org/doc/draft-ietf-storm-iscsi-sam/
> > > >
> > > > iSCSI was originally based on version 2 of the SCSI architecture, SAM-2.
> > > > This draft updates iSCSI to the current version of the SCSI architecture,
> > > > SAM-5, by adding additional features, and a text key to negotiate their
> > > > usage, iSCSIProtocolLevel.  The draft is only about 20 pages - please take
> > > > a look at it from an implementer's standpoint.
> > > >
> > > > Finally, if there's anything you wish the iSCSI RFCs said, or functionality
> > > > that you think should be removed, please say so.
> > > >
> > > > Your comments can be sent directly to the mailing list - [e-mail],
> > > > or can be sent to me.  Please identify yourself as a VMware iSCSI
> > > > implementer in your comments.  The Working Group Last Call on these drafts
> > > > runs through August 21st - please let me know when you think you could
> > > > have comments ready.
> > > >
> > > > Thanks,
> > > > --David
> > > > ----------------------------------------------------
> > > > David L. Black, Distinguished Engineer
> > > > EMC Corporation, 176 South St., Hopkinton, MA  01748
> > > > [phone]             FAX: [phone]
> > > > [e-mail]        Mobile: [phone]
> > > > ----------------------------------------------------
> > >
> _______________________________________________
> storm mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/storm