Re: [DMM] Call for adoption confirmation: draft-seite-dmm-rg-multihoming-02

[Mingui Zhang]

Hi, 

Thanks for letting us know the draft does not intend to define the BBF hybrid 
access arch any more. And yes, I noticed the reference [WT-348] had been 
removed in the 02 version, but "Hybrid Access" is a term defined by BBF WT-348 
so it should be removed from the text to avoid misleading.

> >  As mentioned in Prague and Yokohama,  the use case defined is the
> > broadband use case for hybrid access. Without having proper discussion
> > on the use case it will improper to look for the solution as we MAY
> > not even know the complete architecture issues involved in broadband
> > access hybrid use case which is what the Figure 1 refers too.


Definitely. As the draft intends not to refer to BBF WT-348's "DSL+LTE" 
use-case, Figure 1 should be removed.  

Thanks,
Mingui

> -Original Message-
> From: dmm [mailto:dmm-boun...@ietf.org] On Behalf Of
> pierrick.se...@orange.com
> Sent: Friday, December 04, 2015 5:16 PM
> To: Muley, Praveen V (Praveen); sarik...@ieee.org; Jong-Hyouk Lee
> Cc: dmm
> Subject: Re: [DMM] Call for adoption confirmation:
> draft-seite-dmm-rg-multihoming-02
> 
> Hi,
> 
> I think there is big misunderstood here... this draft does not intend to 
> define
> the BBF hybrid access architecture... not even to address this specific 
> use-case.
> Few months ago, we had a consensus (including AD) to not focus on a specific
> use-case; we thus have removed all references to BBF use-case from this draft.
> 
> As already said, this draft defines protocol extension. How to use this 
> extension
> is part of system design and is out of the scope.
> 
> BTW, PMIP is not a tunneling but a control protocol, underlying tunneling can
> be GRE, IP-in-IP,...
> 
> Pierrick
> 
> > -Message d'origine-
> > De : Muley, Praveen V (Praveen)
> > [mailto:praveen.mu...@alcatel-lucent.com]
> > Envoyé : vendredi 4 décembre 2015 09:45 À : SEITE Pierrick IMT/OLN;
> > sarik...@ieee.org; Jong-Hyouk Lee Cc : dmm Objet : RE: [DMM] Call for
> > adoption confirmation: draft-seite-dmm-rg-multihoming-
> > 02
> >
> > Hi :
> >
> >  As mentioned in Prague and Yokohama,  the use case defined is the
> > broadband use case for hybrid access. Without having proper discussion
> > on the use case it will improper to look for the solution as we MAY
> > not even know the complete architecture issues involved in broadband
> > access hybrid use case which is what the Figure 1 refers too.  So its
> > too pre-mature (oppose) to adopt this as WG document.
> >
> >If the use case different than multihomed RG then please have
> > that network diagram in document and explain properly the use case.
> >
> > If you are saying this solution can be used for multi-homed RG , first
> > of all why would you solve that use case using layers of tunnel as
> > there are some other better solutions which doesn't require tunneling
> > overlay. Given that these RG is very cheap device the performance of
> > these devices goes down dramatically so obviously doesn't address the
> > fundamental requirement of increasing the bandwidth to the end user.
> >
> > Why PMIP even if I have to use tunnel ?  There are many other better
> > tunneling technologies like soft-GRE.
> >
> > One another point is how fast the failure of underlay tunnel on LTE
> > and DSL detected by LMA to avoid blackholing of traffic. It will be
> > good if you can touch base on this in your draft.
> >
> >
> 
> This draft only deals with mPCoAfailure detection is adressed
> 
> >
> > -Praveen
> >
> >
> >
> >
> >
> > -Original Message-
> > From: dmm [mailto:dmm-boun...@ietf.org] On Behalf Of
> > pierrick.se...@orange.com
> > Sent: Wednesday, December 02, 2015 12:22 AM
> > To: sarik...@ieee.org; Jong-Hyouk Lee
> > Cc: dmm
> > Subject: Re: [DMM] Call for adoption confirmation: draft-seite-dmm-rg-
> > multihoming-02
> >
> > Hi Behcet,
> >
> > This extension is for any use-case requiring a MAG to be multihomed.
> > For sure, multihomed RG can be one of them, but there is no reason to
> > restrict MCoA to this use-case.
> >
> > Pierrick
> >
> > > -Message d'origine-
> > > De : dmm [mailto:dmm-boun...@ietf.org] De la part de Behcet Sarikaya
> > > Envoyé : mardi 1 décembre 2015 18:18 À : Jong-Hyouk Lee Cc : dmm
> > > Objet
> > > : Re: [DMM] Call for adoption confirmation:
> > > draft-seite-dmm-rg-multihoming-
> > > 02
> > >
> > > Hi Jong-Hyouk,
> > >
> > > On Mon, Nov 30, 2015 at 10:34 AM, Jong-Hyouk Lee
> > > 
> > > wrote:
> > > > Hi all
> > > >
> > > > I support the adoption of this draft as a WG draft even with the
> > > > concerns pointed by Mingui. This draft has a merit of the
> > > > introduction of the generic protocol extension allowing a
> > > > multihomed MAG
> > >
> > > No, the extension is for the RG, i.e. Residential Gateway which is a
> > > broadband or fixed network element.
> > >
> > >
> > > > to register more than one PCoA
> > > > to the LMA. It is definitely useful for a multihomed environment.
> > >
> > > Why would a MAG 

Re: [DMM] Call for adoption confirmation: draft-seite-dmm-rg-multihoming-02

[Sri Gundavelli (sgundave)]

> but "Hybrid Access" is a term defined by BBF WT-348 so it should be
>removed from the text to avoid misleading.

We will consider this feedback.


> As the draft intends not to refer to BBF WT-348's "DSL+LTE" use-case,
>Figure 1 should be removed.
 

The extension is not tied to any specific access technology and its
perfectly reasonable to refer to any access technology. We will make the
wording clear that its only an example.



Sri





On 12/6/15, 7:29 PM, "dmm on behalf of Mingui Zhang"  wrote:

>Hi, 
>
>Thanks for letting us know the draft does not intend to define the BBF
>hybrid access arch any more. And yes, I noticed the reference [WT-348]
>had been removed in the 02 version, but "Hybrid Access" is a term defined
>by BBF WT-348 so it should be removed from the text to avoid misleading.
>
>> >  As mentioned in Prague and Yokohama,  the use case defined is the
>> > broadband use case for hybrid access. Without having proper discussion
>> > on the use case it will improper to look for the solution as we MAY
>> > not even know the complete architecture issues involved in broadband
>> > access hybrid use case which is what the Figure 1 refers too.
>
>
>Definitely. As the draft intends not to refer to BBF WT-348's "DSL+LTE"
>use-case, Figure 1 should be removed.
>
>Thanks,
>Mingui
>
>> -Original Message-
>> From: dmm [mailto:dmm-boun...@ietf.org] On Behalf Of
>> pierrick.se...@orange.com
>> Sent: Friday, December 04, 2015 5:16 PM
>> To: Muley, Praveen V (Praveen); sarik...@ieee.org; Jong-Hyouk Lee
>> Cc: dmm
>> Subject: Re: [DMM] Call for adoption confirmation:
>> draft-seite-dmm-rg-multihoming-02
>> 
>> Hi,
>> 
>> I think there is big misunderstood here... this draft does not intend
>>to define
>> the BBF hybrid access architecture... not even to address this specific
>>use-case.
>> Few months ago, we had a consensus (including AD) to not focus on a
>>specific
>> use-case; we thus have removed all references to BBF use-case from this
>>draft.
>> 
>> As already said, this draft defines protocol extension. How to use this
>>extension
>> is part of system design and is out of the scope.
>> 
>> BTW, PMIP is not a tunneling but a control protocol, underlying
>>tunneling can
>> be GRE, IP-in-IP,...
>> 
>> Pierrick
>> 
>> > -Message d'origine-
>> > De : Muley, Praveen V (Praveen)
>> > [mailto:praveen.mu...@alcatel-lucent.com]
>> > Envoyé : vendredi 4 décembre 2015 09:45 À : SEITE Pierrick IMT/OLN;
>> > sarik...@ieee.org; Jong-Hyouk Lee Cc : dmm Objet : RE: [DMM] Call for
>> > adoption confirmation: draft-seite-dmm-rg-multihoming-
>> > 02
>> >
>> > Hi :
>> >
>> >  As mentioned in Prague and Yokohama,  the use case defined is the
>> > broadband use case for hybrid access. Without having proper discussion
>> > on the use case it will improper to look for the solution as we MAY
>> > not even know the complete architecture issues involved in broadband
>> > access hybrid use case which is what the Figure 1 refers too.  So its
>> > too pre-mature (oppose) to adopt this as WG document.
>> >
>> >If the use case different than multihomed RG then please have
>> > that network diagram in document and explain properly the use case.
>> >
>> > If you are saying this solution can be used for multi-homed RG , first
>> > of all why would you solve that use case using layers of tunnel as
>> > there are some other better solutions which doesn't require tunneling
>> > overlay. Given that these RG is very cheap device the performance of
>> > these devices goes down dramatically so obviously doesn't address the
>> > fundamental requirement of increasing the bandwidth to the end user.
>> >
>> > Why PMIP even if I have to use tunnel ?  There are many other better
>> > tunneling technologies like soft-GRE.
>> >
>> > One another point is how fast the failure of underlay tunnel on LTE
>> > and DSL detected by LMA to avoid blackholing of traffic. It will be
>> > good if you can touch base on this in your draft.
>> >
>> >
>> 
>> This draft only deals with mPCoAfailure detection is adressed
>> 
>> >
>> > -Praveen
>> >
>> >
>> >
>> >
>> >
>> > -Original Message-
>> > From: dmm [mailto:dmm-boun...@ietf.org] On Behalf Of
>> > pierrick.se...@orange.com
>> > Sent: Wednesday, December 02, 2015 12:22 AM
>> > To: sarik...@ieee.org; Jong-Hyouk Lee
>> > Cc: dmm
>> > Subject: Re: [DMM] Call for adoption confirmation: draft-seite-dmm-rg-
>> > multihoming-02
>> >
>> > Hi Behcet,
>> >
>> > This extension is for any use-case requiring a MAG to be multihomed.
>> > For sure, multihomed RG can be one of them, but there is no reason to
>> > restrict MCoA to this use-case.
>> >
>> > Pierrick
>> >
>> > > -Message d'origine-
>> > > De : dmm [mailto:dmm-boun...@ietf.org] De la part de Behcet Sarikaya
>> > > Envoyé : mardi 1 décembre 2015 18:18 À : Jong-Hyouk Lee Cc : dmm
>> > > Objet
>> > > : Re: [DMM] Call for adoption confirmation:
>> > >