[Idr] Re: draft-decraene-idr-nlri-error-handling-00.txt

[Robert Raszuk <[e-mail]>]

I mean when you can't parse just one path_id and rest are all cool why
would you consider entire MP_REACH as toxic and treat-as-withdraw all paths
for all NLRIs ?

On Fri, Oct 17, 2025 at 9:40 PM John Scudder <[e-mail]> wrote:

> On Oct 17, 2025, at 3:23 PM, Robert Raszuk <[e-mail]> wrote:
>
> With ADD-PATHs MP_UNREACH is quite different then MP_REACH.
>
>
> What do you mean? In both of them, the NLRI is just the catenation of the
> Path ID and the NLRI as it would be without RFC 7911. From Section 3:
>
>    In order to carry the Path Identifier in an UPDATE message, the NLRI
>    encoding MUST be extended by prepending the Path Identifier field,
>    which is of four octets.
>
> RFC 7911 does not even talk about MP_REACH or MP_UNREACH.
>
> Is one match of NLRI found in TAW enough irrespective of how many matches
> for the NLRI MP_REACH with extended NLRI encoding would/should contain ?
>
>
> I don’t understand this question.
>
> —John
>
>