Re: I-D Action: draft-ietf-idr-deprecate-as-set-confed-set-02.txt

[resending...   I meant to cc IDR list]

Jeff,

>This gets to be an interesting bit of code verification since it means that
>some implementations may have stripped the 2-byte version of sets and the
>4-byte stuff is still present in the AS4_PATH.

IMO, this situation will not arise because there no "stripping" of AS_SET.
There is only Treat-as-withdraw.

Thanks.

Sriram

==================================================
[Ondrej]
> Hi
>
> Just added a patch [*] to BIRD to optionally reject AS_SETs (and AS_CONFED_SETs).
>
> One thing that is not discussed in the draft is handling of AS_SETs in AS4_PATH.
> It would make sense to also do treat-as-withdraw, but handling of inappropriate
> AS_CONFED_* segments is already different in AS_PATH and AS4_PATH.

[Jeff]
This is a very good point.  The related normative updates vs. RFC 4893 are
now required.

This gets to be an interesting bit of code verification since it means that
some implementations may have stripped the 2-byte version of sets and the
4-byte stuff is still present in the AS4_PATH.

Then again, the spec was always incomplete with all of the crazy edge cases
that could happen.

_______________________________________________
Idr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/idr