hi Ian,

2013/2/8 <[email protected]>

> HI Maoke,
>
> Comments in line.
>
> Cheers,
> Ian
>
>
> Deutsche Telekom AG
> Group Technology
> Ian Farrer
> IP Engineering
> Landgrabenweg 151, 53227 Bonn, Germany
> +49 228 93638046 (Phone)
> +49 170 4557418 (Mobile)
> E-Mail: [email protected]
>
> Life is for sharing.
>
> Deutsche Telekom AG
> Supervisory Board: Prof. Dr. Ulrich Lehner (Chairman)
> Board of Management: René Obermann (Chairman),
> Reinhard Clemens, Niek Jan van Damme, Timotheus Höttges,
> Dr. Thomas Kremer, Claudia Nemat, Prof. Dr. Marion Schick
> Commercial register: Amtsgericht Bonn HRB 6794
> Registered office: Bonn
> VAT identification no. DE 123475223
>
> *Big changes start small **–** conserve resources by not printing every
> e-mail.*
>
> From: Maoke <[email protected]>
> Date: Wednesday, 6 February 2013 02:58
> To: Ian Farrer <[email protected]>
> Cc: "[email protected]" <[email protected]>, "
> [email protected]" <[email protected]>, "[email protected]" <
> [email protected]>, "[email protected]" <[email protected]>
>
> Subject: Re: [Softwires] Call for adoption of
> draft-bfmk-softwire-unified-cpe-02
>
>
>
> 2013/2/5 <[email protected]>
>
>> Hi Maoke,
>>
>> Thanks for your response. Comments inline.
>>
>> Cheers,
>> Ian
>>
>> From: Maoke <[email protected]>
>> Date: Tuesday, 5 February 2013 06:54
>> To: Suresh Krishnan <[email protected]>
>> Cc: Softwires WG <[email protected]>, Yong Cui <[email protected]>,
>> Ralph Droms <[email protected]>
>> Subject: Re: [Softwires] Call for adoption of
>> draft-bfmk-softwire-unified-cpe-02
>>
>>
>> hi Suresh and all,
>>
>> personally i think this work is still premature for the WG adoption.
>> reasons are below:
>>
>> 1. current draft makes a logic for the CPE to decide its mode according
>> to what information it received from DHCP options. however, DHCP options
>> are not the only available provision mean. what if my CPE get MAP
>> information through DHCP, while lw4over6 information from PCP, while
>> another MAP domain's provision through manually configuration? when such
>> kind of basic questions not yet fully discussed, i doubt it is suitable to
>> adopt this work at so early stage.
>>
>> [ian] The intention for the draft is to describe using the presence (or
>> absence) of configuration parameters so that a CPE can work out which mode
>> to configure. The DHCP section is there to show how this would work if you
>> were using DHCP as the config mechanism, but it doesn't
>>
>>
> [maoke] the draft states "e.g." for the DHCP configuration firstly but
> later the total discussion is based on only DHCP. i doubt this can be a
> good direction for the further approach.
>
> [ian] That's not the intention. Only section 3.3 discusses DHCP in depth
> to show how this works. The implementation is given as a should req. to be
> in line with the DHCP provisioning should requirements in both MAP and
> lw4o6. Point taken however – I'll try and make the rest of the document
> (e.g. Except sec 3.3) less DHCP focused.
>

[maoke] in contrast, i don't think DHCP MUST be the only way for
provisioning. i'd prefer the "e.g." rather than limiting the further
mechanisms DHCP-focused.


>
>
>> 2. some important issues are not yet comprehensively discussed, or almost
>> missing, in the current draft. e.g. (but not limited to),
>>     - what is the correct NAPT source port overlapping behaviour for the
>> unified CPE?
>>     - what is the correct fragment/reassemble behaviour for the unified
>> CPE?
>>
>> [ian] I agree that these are problems that need to be addressed across
>> both MAP and LW, but both of these problems need solutions that are aligned
>> at both ends of the softwire, not just the CPE.
>>
>
> [maoke] NAPT issue is totally CPE issue. i don't agree it is a problem for
> the *both* ends. fragmentation issue involves both sides, but CPE behaviour
> impacting the anycast BR/AFTR performance is a well-known issue in the
> community. even BR/AFTR has any change, the CPE must contain a work on this
> issue.
>
> [ian] If there are overlapping source ports, then this is also a problem
> for the concentrator. If multiple initiators share the same v4 destination
> address / source ports, on what basis can the initiator forward ingress
> traffic?
>

[maoke] we have had at least 3 alternative strategies regarding source port
overlapping and are comparing their pros/cons. practical usage has proved
the availability to some extent.


>
>
>> [ian] For the fragmentation question, Ole proposed this as being a
>> suitable task to tackle as it is common to all solutions. I would propose
>> that this is how it is tackled, and then any affected drafts are updated
>> based on the outcome.
>>
>> as currently there is no MAP members joins this draft, i am worrying that
>> the draft has fully reflected the concern from the MAP side.
>>
>>
>> [ian] As there haven't been any comments on the draft for some time from
>> anyone involved in the WG, I'm hoping that this means that everyone is
>> happy with the content!
>>
>
> [maoke] happy with the content at current stage = acknowledge to the
> authors' effort != happy with the content at the level of WG adoption. ;-)
> for the time being...
>
> [ian] Let's see what the outcome of the adoption call is then.
>

[maoke] i insist to propose considering the problems: 1) if DHCP is the
only way that the unified CPE can do with? 2) if participating multiple
domain is/cannot definitely be supported? 3) how the CPE support source
port overlapping and fragmentation with multiple domain exits? i
contributed these concerns to the community -- it is enough for me. if the
draft is less qualified with my criteria, i just do not use it.

as to the adoption, i have expressed my comments. the result is not in my
concern.

thanks and regards,
maoke


>
>
> - maoke
>
>
>>
>> thanks and regards,
>> maoke
>>
>> 2013/2/5 Suresh Krishnan <[email protected]>
>>
>>> Hi all,
>>>   This draft was a result of the discussion initiated at the softwire
>>> meeting in Atlanta to attempt to come up with a unified CPE
>>> specification that can work with both MAP and lw4o6. This call is being
>>> initiated to determine whether there is WG consensus towards adoption of
>>> draft-bfmk-softwire-unified-cpe-02 as a softwire WG draft. Please state
>>> whether or not you're in favor of the adoption by replying to this
>>> email. If you are not in favor, please also state your objections in
>>> your response. This adoption call will complete on 2013-02-18.
>>>
>>> Regards
>>> Suresh & Yong
>>>
>>>
>>> _______________________________________________
>>> Softwires mailing list
>>> [email protected]
>>> https://www.ietf.org/mailman/listinfo/softwires
>>>
>>
>>
>
_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires

Reply via email to