Re: [spring] SRH scratch space (was Re: Question about SRv6 Insert function)

[Robert Raszuk <[e-mail]>]

Thank you.

Yes this draft can work with controller, but can also work without one.

Please observe that I am (re)using BGP extension to propagate control plane
information which can originate from any network element not necessarily a
central controller. Nevertheless using it with conjunction of some oracle
to optimize global network TE or steer some packets towards selected SFCs
is by all means a sound approach.

Kind regards,
Robert

On Thu, Dec 12, 2019 at 7:54 PM Gyan Mishra <[e-mail]> wrote:

>
>
> On Wed, Dec 11, 2019 at 3:37 AM Robert Raszuk <[e-mail]> wrote:
>
>> Hi Brian,
>>
>> > Each SR node (Segment Endpoint) is effectively applying new IPv6 encap
>>> so is already doing an insertion of new SRH
>>>
>>> That isn't "insertion" in the sense of draft-voyer.
>>
>>
>> Correct. But we are not discussing draft-voyer here.
>>
>> I think recent mail threads prove that it is much better to discuss one
>> topic at a time rather then mix three different and pretty orthogonal
>> "issues" interchangeably.
>>
>>
>>>  . It's prepending an extra layer of encapsulation, which is indeed just
>>> fine and I don't think anyone here is objecting to it. The spring draft
>>> currently uses (IMHO) imprecise language, even in its pseudo-code, but if
>>> all it's doing is describing successive layers of encapsulation that's fine
>>> too. Wasted bytes perhaps, but that isn't an IETF problem.
>>
>>
>> Great that we agree.
>>
>>
>>> Whether prepending new headers, each with their own SRH, is the best way
>>> of doing service-based traffic engineering is another question. Having just
>>> reviewed draft-ietf-bess-nsh-bgp-control-plane for Gen-ART, I do believe
>>> there's more than one possible approach.
>>>
>>
>> 100% agreed. In fact as you may have seen I have another non data plane,
>> but control plane proposal posted:
>> https://tools.ietf.org/html/draft-raszuk-teas-ip-te-np-00
>>
>> Kind regards,
>> R.
>>
>
>   Robert
>
>     Your ip-te teas WG draft sounds like a fine solution and maybe can be
> used in conjunction with Gen-ART sfc draft in centralized SDN controller
> based architecture.  So with draft is pulls the control plane function or
> Ti-LFA and SR-TE and/or binding SID policies into control plane nodes which
> could be the PEs eliminating SRH state in the P core.  So this in essence
> using this feature we would save on overhead bytes with the additional 6in6
> encap plus SRH state.
>
> Warm regards,
>
> Gyan
>
>>
>> --------------------------------------------------------------------
>> IETF IPv6 working group mailing list
>> [e-mail]
>> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
>> --------------------------------------------------------------------
>>
> --
>
> Gyan S. Mishra
>
> IT Network Engineering & Technology
>
> Verizon Communications Inc. (VZ)
>
> 13101 Columbia Pike FDC1 3rd Floor
>
> Silver Spring, MD 20904
>
> United States
>
> Phone: [phone]
>
> Email: [e-mail]
>
> www.linkedin.com/in/networking-technologies-consultant
>
>