Re: [Idr] WG Adoption call for draft-wang-idr-rd-orf-05.txt (2/4/2021 to 2/18/2021)

[Robert Raszuk <[e-mail]>]

Sorry typo

s/PEs which import routes do not understand sender's ORF./PEs which import
routes do not understand sender's RDs/

On Fri, Feb 12, 2021 at 4:09 PM Robert Raszuk <[e-mail]> wrote:

> The receiver device, such as the access router, may have weak performance,
>> while the sender device, such as the RR, may have strong performance.
>>
>
> Yes., That is precisely why we have  https://tools.ietf.org/html/rfc4684
>
> The ORF is a tool and does not exist independently. The deployment problem
>> described here occurs only when the RD-ORF mode is used.
>>
>> Actually, the ORF can be used together with other ORFs, such as Prefixed
>> ORF.
>>
>> RD-ORF provides a more efficient filtering method, making solution
>> deployment more flexible.
>>
>>
>>
>> In fact, most vendors' devices support rd-filter, but only support the
>> configuration.
>>
>>
>>
>> In addition, if the RR decides to push the RD-ORF to its route source
>> only when PE1 and PE3 do not want that RD.
>>
>
> See this is one more time indication that you have serious gap in
> understanding L3VPNs.
>
> PEs which import routes do not understand sender's ORF. They take routes
> and based on RT make a local copy to importing VRF.
>
> In RFC4364 RTs are used to build arbitrary VPN topologies not RDs.
>
> Thx
>
>
>