Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity



John-

I applaud that you are trying to get a leg up on the next generation high speed project.
That is a good thing
The point that decisions would/could be made before we get to the study group phase is entirely valid and would to some extent move decisions out of 802.3 to another venue. E.g. choosing a speed is easily done by whoever makes the CFI.  RTPGE is an example of this.
I believe you think this is a good thing.  I won't put forth any opinion at the moment

What everyone should realize is that there is no guarantee that an entity membership Industry Connections group will end up acting as a well-behaved sub-group of 802.3.  For that matter it might end up with no allegiance to 802 whatsoever.

The Industry Connections process was designed by the IEEE to nurture and produce entity based standards projects.
We would be trying to use the process for another purpose.

According to the SA's web page: "ABOUT THE PROGRAM"
    REF: http://standards.ieee.org/develop/indconn/index.html
"The IEEE-SA Industry Connections (IC) program helps ... like-minded organizations."
and
"Industry Connections empowers groups ... to produce "fast-track" materials, such as, proposals for standards"  (i.e. PARS)

I'm not at all sure that Industry Connections is an appropriate and safe forum for an activity whose purpose is to support future work in 802.3.

Respectfully,

    Geoff Thompson

Geoffrey O. Thompson
GraCaSI S.A.
Mountain View, CA 94043
<thompson@xxxxxxxx>



On 286//12 2:43 AM, John D'Ambrosia wrote:

All

Let me respond to some of the issues that have been raised in this thread.

 

First of all, while some may feel things are more definitive of what can and can’t be done, what will or won’t be done, my take on proposing this IC has been to not factor in any assumed outcomes.  I have read these response interpreting people’s words as their opinions, and frankly, IMO, it illustrates the need for this open industry discussion. 

 

Next, the IC has been proposed in such a manner that it is not intended to be disruptive to the defined actual project process.  It is not proposing changing any rules where an IC would have to be done or completed prior to an actual CFI. 

 

All efforts I have been involved with in the past have never been closed or private.  The IC is merely another available tool for building consensus towards a CFI by providing a forum for industry discussion where awareness and participation of all potential stakeholders can be facilitated through the IEEE, which I believe all would agree, would help enable even broader reach and hence greater participation.  The only proposed deliverables is the historical records, i.e. meeting notes & presentations.  There is no decision outcome, though I think the forum would allow us the opportunity to do non-binding straw polls to get an assessment of the interest level.  I believe it is fair to say that a perceived outcome is that this IC will be helpful to a potential future CFI.

 

The rate question is perhaps one of the most explosive issues related to this whole effort, and it suggests that we review our rules.  Per the IEEE 802.3 Operating Rules, Section 4.2 Formation –

 

“A SG is formed when enough interest has been identified for a particular area of study within the scope of the WG. The first step in the process is a call for interest.”

 

So, in regards to the rate question, while Steve used the phrasing “make the rate decision during the CFI”, another way one might phrase the intent of his statement is that the IC would enable the group to evaluate the consensus of chooing 400G as the proposed area of study.  (Given my prior experience chairing the 2006-2007 HSSG, I hope some will understand why I can see the value if a choice were made).  Traditionally, however, my recollection is that no motion for a new speed project has ever stated the rate choice in the chartering motion.  Nonetheless, I personally wouldn’t see such a motion to be prohibited by the rules, but one cannot look at this issue with tunnel vision.  The consensus of the group, IMO, will have an impact on what motion might be decided to be proposed, how that motion might be treated on the floor during the WG session, i.e. amendments, and ultimately whether the motion passes.  I can share from my prior experience that there were many individuals who felt I should have included 100G in the chartering motion for the last HSSG.  Based on on-going conversations related to those I spoke with however, my own personal determination was that the inclusion of 100G would have been disruptive from my ultimate goal at that time, however, which was getting the Study Group formed.  And while some have very strong opinions regarding the current rate debate, which they have expressed, I can personally say, that in organizing this IC, I have heard strong opinions supporting differing points of views on the rate topicsubject. (Another example I would argue for this IC!)  And thus, I have proposed the IC in such a way that no decisions are implied, the IC merely provides a forum for the discussion and consensus building. 

 

I hope this provides the clarification that some individuals are looking for.

 

Regards,

 

John

 

 

 

 

From: Brad Booth [mailto:Brad_Booth@xxxxxxxx]
Sent: Wednesday, June 27, 2012 6:32 PM
To: STDS-802-3-DIALOG@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity

 

Steve,

 

That’s for the explanation. It’s definitely a good level set for those that haven’t been through previous HSSG incarnations or who have not participated in call-for-interest (CFI) developments.

 

One aspect you mentioned (and I assume was a typo) is the statement, “make the rate decision during the CFI.” As you and I (and many others) know, the only decision that occurs with a CFI is whether or not to form a study group. The CFI is made Monday afternoon at the 802.3 opening session and voted upon at the closing session Thursday afternoon.

 

What is not mentioned in the email is whether or not the outcome of the industry connections (IC) is going to be a CFI. For example, if the intended outcome of the IC was stated as, “we will do a CFI for speed X” then that has a different impact than, “we may do a CFI for the next speed.” If the IC focus is to select the speed for the CFI and drive a CFI development, then that makes sense for companies interested to invest time and resources. If the IC effort is not going to select a speed or lead to a CFI, then other than networking with colleagues at a set place and time, this would result in companies investing time and resources with nothing to show for it.

 

Respectfully, it was not 100% clear. All I’m asking for is clarification.

 

Thanks,

Brad

 

 

From: Trowbridge, Stephen J (Steve) [mailto:steve.trowbridge@xxxxxxxxxxxxxxxxxx]
Sent: Wednesday, June 27, 2012 4:15 PM
To: Booth, Brad; STDS-802-3-DIALOG@xxxxxxxxxxxxxxxxx
Subject: RE: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity

 

Hi Brad and all,

I would like to speak in favor of this IC effort – like many of the others, I find that a lot of folks that I speak to, both inside of my company and at industry events, standards meetings and conferences are quite eager to see something get started on the “next rate”, and the level of interest is something that I think justifies an activity to meet the need.

 

You have observed, quite correctly, that we haven’t done it this way in the past several HSSGs in 802.3 history, so it is worth discussing what makes this different. I think there are two things that justify a different approach:

 

The first is that “pre-CFI” activities have generally been private efforts among a (small) group of like-minded participants who do the work to justify the project. If they meet during 802.3 meetings, they do so informally, as a “flash mob”, using one of the meeting rooms but not advertising it on the schedule. Given the large number of participants and companies who are interested in this, however, it makes more sense to do it as an open process, where anyone interested can join – to put the room on the announced schedule of sessions at 802.3 meetings, and to have an open email list that participants can join.

 

The second reason, which became quite apparent during 802.3ba, is that Ethernet is part of an ecosystem and does not stand by itself. Particularly at these very high rates, we are beyond the point where anyone will use such an interface anytime soon to connect to an end station, so these interfaces are generally carrying very large aggregates. For the part of the network where the next rate will play, you need to be able to carry that new rate of Ethernet over OTN. It could be that the IC effort puts together enough information to know that we need a faster rate without knowing what that next rate should be, but it is also be the case that what the IC learns will give good information for what the next rate should be. If this occurs, it would benefit the industry to make that decision early (e.g., to make the rate decision during the CFI), since that would enable IEEE and ITU-T work to proceed independently for each organization to work on developing aligned standards in parallel.

Regards,
Steve

 

From: Brad Booth [mailto:Brad_Booth@xxxxxxxx]
Sent: Tuesday, June 26, 2012 10:00 AM
To: STDS-802-3-DIALOG@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity

 

I guess I’ll be the first of those not participating in the BWA tutorial to speak relative to this new IC effort.

 

The question is being raised in the industry about what is the next speed of Ethernet and when will that occur. Personally, I have been asked that question ever since we started the previous Higher Speed Study Group (HSSG) that resulted in IEEE Std. 802.3ba-2010. The Bandwidth Assessment Ad Hoc helped explore and shed light on the drivers and growth potential that exists. This was a great industry connections effort that the IEEE 802.3 Working Group gets to put its stamp on.

 

What is harder to understand is how an industry connections effort leads into the potential development of a call-for-interest (CFI). During my 15 years in 802.3, speed increments CFIs have always been referred to as HSSG as the study group would pick the next higher speed. Is the industry connections effort looking to build consensus on what the next speed should be for the CFI and eliminate that selection process from the study group phase of the project? Or, will the next speed consensus built within the industry connection effort be irrelevant as the HSSG will still be in charge of selection the next speed?

 

I’m in favor of developing the next HSSG when the time is right. I also believe that the 802.3 Working Group should understand the process of how an industry connections activity feeds into a CFI. This is new territory for 802.3 and it would seem prudent to make sure there are no false expectations or misperceptions.

 

Sincerely,

Brad

 

 

Brad Booth

Director, Network Architecture

Social Media and Community Professional

Dell | Office of the CTO, Networking

brad_booth@xxxxxxxx

office +1-512-723-7106, mobile +1-503-830-9366

 

 

 

From: Peter Stassar [mailto:Peter.Stassar@xxxxxxxxxx]
Sent: Tuesday, June 26, 2012 9:14 AM
To: STDS-802-3-DIALOG@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity

 

Hi all,

 

I would like to express my full support for this IC activity in line with previous support messages from Pete Anslow and Mark Nowell.

 

Kind regards,

 

Peter

 

Peter Stassar

Technical Director

Huawei Technologies Ltd

European Research Center

Karspeldreef 4, 1101CJ Amsterdam

The Netherlands

Tel: +31 20 4300 832

Mob: +31 6 21146286

--------------------------------------------------------

    To: STDS-802-3-DIALOG@xxxxxxxxxxxxxxxxx

    Subject: Re: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity

    From: "Mark Nowell (mnowell)" <mnowell@xxxxxxxxx>

    Date: Tue, 26 Jun 2012 09:21:54 -0400

    Delivered-to: mhonarc@xxxxxxxxxxxxxxxx

    In-reply-to: <29C931202891DE4ABE913168E7AC43ECA85091FE@xxxxxxxxxxxxxxxxxxxxxx>

    List-help: <https://listserv.ieee.org/cgi-bin/wa?LIST=STDS-802-3-DIALOG>, <mailto:LISTSERV@xxxxxxxxxxxxxxxxx?body=INFO%20STDS-802-3-DIALOG>

    List-owner: <mailto:STDS-802-3-DIALOG-request@xxxxxxxxxxxxxxxxx>

    List-subscribe: <mailto:STDS-802-3-DIALOG-subscribe-request@xxxxxxxxxxxxxxxxx>

    List-unsubscribe: <mailto:STDS-802-3-DIALOG-unsubscribe-request@xxxxxxxxxxxxxxxxx>

    References: <256C0B6368B8E84AA888EF260D95C8654E9F6527@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <29C931202891DE4ABE913168E7AC43ECA85091FE@xxxxxxxxxxxxxxxxxxxxxx>

    Reply-to: "Mark Nowell (mnowell)" <mnowell@xxxxxxxxx>

    Thread-index: Ac1TnqhgjNH4w5wTSnu93iai8wcVrg==

    Thread-topic: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity

 

I agree with Pete. We have been seeing growing interest from numerous network operators covering SP, MSO and Data Center spaces, to have clarity on when a next higher speed of Ethernet would be available.  Concern is that it will be available in time for their projected network needs. 

 

The sustained network bandwidth growth, driven largely by the increase in users and the increase in user consumption (such as  video) is challenging network's ability to scale.  Built into this need for a speed step is also the challenge of ensuring a cost effective solution. 

 

At the last interim meeting a number of interested people got together for an ad-hoc discussion around this topic.  At the close we took a straw poll to gauge interest in starting things moving towards a next HSSG project. Results indicated that there was strong interest in kicking this off.

 

In light of this, I support the

 

On Jun 26, 2012, at 7:01 AM, "Anslow, Peter" <panslow@xxxxxxxxx> wrote:

 

> John,

> Thanks for initiating this new 'Higher Speed Ethernet' Industry Connections activity.

> I have seen a tremendous amount of interest in higher rates of Ethernet at conferences and workshops since 802.3ba was completed, so I think that it is a good idea to start an activity within 802.3 aimed at gathering consensus on a way forward towards a future CFI on this topic.

> Regards,

> Pete Anslow | Senior Standards Advisor

> 43-51 Worship Street | London, EC2A 2DX, UK

> Direct +44 2070 125535 |

> _____________________________________________

> From: Law, David [mailto:dlaw@xxxxxx]

> Sent: 22 June 2012 08:44

> To: STDS-802-3-DIALOG@xxxxxxxxxxxxxxxxx

> Subject: [802.3_DIALOG] New 'Higher Speed Ethernet Consensus' Industry Connections activity

> Dear Colleagues,

> I have received a request from John D'Ambrosia to forward the following message to the IEEE 802.3 Working Group reflector. As the IEEE 802.3 Working Group reflector is for announcements only I have also CCed the IEEE 802.3 general discussion reflector. Please therefore direct all discussions to the IEEE 802.3 general discussion reflector <stds-802-3-dilaog@xxxxxxxx>

> The IEEE 802.3 general discussion reflector is an opt-in reflector so if you wish to discuss this issue please subscribe to the IEEE 802.3 general discussion reflector, if you have not already done so, at the URL <http://www.ieee802.org/3/dialog_reflector.html>, or follow the discussion on the reflector email archive at the URL <http://www.ieee802.org/3/email_dialog/thrd1.html>.

> Best regards,

>   David

> ------

> Dear Members of the IEEE 802.3 Working Group,

> Per the July Plenary Meeting Announcement, the Ethernet Bandwidth Assessment Ad Hoc will be seeking Working Group approval of its report. In addition, I will be working with Pete Anslow, Mark Nowell, Scott Kipp, and Peter Stassar to put on a tutorial at the July Plenary <See http://www.ieee802.org/802_tutorials/2012-07/802-0712-TutorialFlyer-V1.pdf>.

> At the IEEE 802.3 May Interim I spoke at most of the Task Force and Study Group meetings (all but MIBs and Maintenance) regarding the future after the IEEE 802.3 BWA. I have been working with a group of approximately 60 individuals on a new 'Higher Speed Ethernet' Industry Connections activity.  The proposed scope of the activity would be to focus on building consensus related to the next speed of Ethernet for wireline applications, which will be used for the evaluation and possible development of a Call-For-Interest for the next Higher Speed Study Group. Please note that unlike the prior BWA activity, the output of this effort would be the historical records, i.e. meeting notes & presentations, of the meetings held by this Industry Connections activity.

> In the interest of full disclosure, an Industry Connections activity Initiation Document (ICAID) was submitted to the Industry Connections Committee (ICCom) for conditional approval, conditional on IEEE 802 EC endorsement in July, at its June 8th meeting. However, ICCom postponed recommending it approval until it has been endorsed by IEEE 802 EC (the sponsoring group), and it was indicated that ICCom will schedule a teleconference the Monday after the plenary, which could then be submitted for a 10 day standards board email ballot.

> Unfortunately, this will have an impact on the formative activities related to the anticipated first meeting of this new IC Activity, as well as the accompanying workshop, where the BWA findings and next steps can be discussed.

> The other option, which I have chosen to pursue, is to forward my request to the IEEE 802 EC for endorsement prior to the July Plenary. This will enable ICCom's approval recommendation sooner, as well as the final approval of the Standards Board.

> I had already requested time at the July opening agenda to give the IEEE 802.3 WG an update, but given the timing of things, wanted to send this email out, so that the members of the WG were aware of the status of my activities.

> Please feel free to contact me if you have any questions.

> Regards,

> John D'Ambrosia