Re: [mpls] Way two progress two mldp draft with an technical overlap

[Loa Andersson <[e-mail]>]

On 2013-12-13 00:51, Yakov Rekhter wrote:
> Loa,
>
>> Yacov,
>>
>> I can't see that this flies! In fact it is counter to the wg chair
>> proposal. We intended to keep the overlap as small as possible,
>> specify a function in the document that needs the function. We did
>> not intend to increase the overlap, but keep it as small as possible.
>>
>> You have this already neatly specified in draft-rekhter- let it stay
>> there.
>
> I am fine with keeping the encoding and the procedures for the two
> new mLDP TLVs: Transit IPv4 Shared Tree TLV, and Transit IPv6 Shared
> Tree TLV in draft-rekhter-mpls-pim-sm-over-mldp.
>
> However, I have a question on the following:
>
>    2.  Assuming that the drafts are adopted, complete
>    draft-wijnands-mpls-in-band-wildcard-encoding as the
>    normative protocol specification of the piece within the
>    overlap.
>
> What do you define as "the overlap" ?
>
> Yakov.

Yakov and Ice,

I'm sorry that I left the two draft hanging in a no mans land for
such a long time, other than traveling, being busy with other working
group issues, and having to do a couple of restarts there are no excuse.

Yakov,

It is my understanding that the overlap is summed up in the following
text from draft-rekhter:

   "This document also identifies the deployment scenarios where BGP
    Source Active auto-discovery routes will not be used."

More specifically, the overlap is the case where the service provider
has provisioned the network in such a way that the RP for a particular
group G is always between the receivers and the sources.  If the
network is provisioned that way, the ingress PE for (S,G) is always
the same as the ingress PE for the RP, so the SA A-D routes are never
needed.

Draft-rekhter should be scoped to the case where the network is not
known to be provisioned in this way.

/Loa
mpls wg co-chair


>
>
>>
>> /Loa
>>
>>
>>
>> On 2013-12-06 22:05, Yakov Rekhter wrote:
>>> Loa,
>>>
>>>> Working Group,
>>>>
>>>> It has been pointed out that there is overlap between
>>>> draft-wijnands-mpls-mldp-in-band-wildcard-encoding and
>>>> draft-rekhter-mpls-pim-sm-over-mldp.  Within the overlap
>>>> there is a critical piece of the protocol specification.
>>>> Having this specified in two places is likely to result
>>>> in non-interoperable implementations.
>>>>
>>>> The working group chairs have discussed the overlap and
>>>> propose the following as a means of moving forward.
>>>>
>>>> 1.  Issue a single poll to adopt both documents together as
>>>> working group documents
>>>>
>>>> 2.  Assuming that the drafts are adopted, complete
>>>> draft-wijnands-mpls-in-band-wildcard-encoding as the
>>>> normative protocol specification of the piece within the
>>>> overlap.
>>>>
>>>> 3.  Where mechanisms in draft-wijnanads are needed in
>>>> draft-rekhter-mpls-pim-sm-over-mldp have the latter document
>>>> reference the necessary sections of the former document.
>>>
>>> This would be fine with me, *provided* that
>>> draft-wijnands-mpls-in-band-wildcard-encoding will add the encoding
>>> of two new mLDP TLVs: Transit IPv4 Shared Tree TLV, and Transit
>>> IPv6 Shared Tree TLV, as these two TLVs are required by the mechanisms
>>> defined in draft-rekhter-mpls-pim-sm-over-mldp.
>>>
>>> Yakov.
>>>
>>
>> --
>>
>>
>> Loa Andersson                        email: [e-mail]
>> Senior MPLS Expert                          [e-mail]
>> Huawei Technologies (consultant)     phone: [phone]
>>
>

-- 


Loa Andersson                        email: [e-mail]
Senior MPLS Expert                          [e-mail]
Huawei Technologies (consultant)     phone: [phone]