Ben,

Perhaps, although AFAIK the three values defined in RFC 4364 have not been 
supplemented since it was published in 2006.

Yours Irrespectively,

John


Juniper Business Use Only

> -----Original Message-----
> From: Benjamin Kaduk <[email protected]>
> Sent: Thursday, August 20, 2020 9:49 AM
> To: John E Drake <[email protected]>
> Cc: Adrian Farrel <[email protected]>; 'The IESG' <[email protected]>;
> [email protected]; [email protected]; bess-
> [email protected]; [email protected]
> Subject: Re: [bess] Benjamin Kaduk's Discuss on 
> draft-ietf-bess-nsh-bgp-control-
> plane-15: (with DISCUSS and COMMENT)
> 
> [External Email. Be cautious of content]
> 
> 
> Hi John,
> 
> On Thu, Aug 20, 2020 at 01:43:34PM +0000, John E Drake wrote:
> > Ben,
> >
> > An RD is encoded using the same format as an extended community but it isn't
> an extended community.  Rather, it is actually part of the NLRI.  The first 
> octet is
> always zero whereas the first octet of an SFIR Pool Identifier extended
> community will always be non-zero (TBD6).
> 
> Thanks for helping clarify.  I'm not seeing what guarantees that the first 
> octet of
> the RD is always zero going forward, though -- I see
> https://urldefense.com/v3/__https://www.iana.org/assignments/route-
> distinguisher-types/route-distinguisher-types.xhtml__;!!NEt6yMaO-
> gk!XaVl3FGXv5cbCKTxNwk3iDLsewVmqJqBOHl8SCveaWdUk0BtkI2cbqt92LXsa5E
> $
> that lists values 3-65534 as "Unassigned".  If we are going to rely on 
> non-overlap
> between the extended community types used for this document and RD types in
> general, don't we need to mark some values in one (or both!) registry as
> Reserved with a reference to [this-document]?
> 
> -Ben

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

Reply via email to