Re: [Idr] Review of draft-ietf-large-community-06.txt

[Job Snijders <[e-mail]>]

On Fri, Nov 04, 2016 at 09:55:49PM +0000, Jakob Heitz (jheitz) wrote:
> All good, except one. You added:
> 
>    There is no significance to the order
>    ...
>    of the BGP Large Communities attributes if two or more such
>    attributes are encoded in the BGP path attribute payload of a single
>    BGP Update message.
> 
> This is in conflict with RFC 7606:
> 
>        If any other attribute (whether
>        recognized or unrecognized) appears more than once in an UPDATE
>        message, then all the occurrences of the attribute other than the
>        first one SHALL be discarded and the UPDATE message will continue
>        to be processed.
> 
> I would leave out any mention of multiple attributes.  What this is
> meant to say is that the order of community values within the
> attribute is irrelevant. I put that in, because it has caused
> confusion in the past. I was thinking to add that "a router may
> reorder the community values when forwarding to the next router",
> because this point was the actual source of confusion in more than one
> instance.

Good catch, Ignas provided text to fix it. Attached are rfcdiff & txt.

Can you double check if that confirms your understanding too with regard
to duplicate attributes & duplicate attribute values.

Kind regards,

Job