BGPTUNNEL [RE: Work on MPLS and tunneling approaches in scenarios]

[Pekka Savola <pekkas@netcore.fi>]

On Tue, 2 Mar 2004, Francois Le Faucheur (flefauch) wrote:
> "bgp-tunnel" is generic and specifies several flavors of how to
> interconnect IPv6 island over an IPv4 cloud with BGP (eg. "MP-BGP over
> IPv4"  as well as "MP-BGP over IPv6" , eg. "Tunneling over IPv4/GRE
> tunnels" as well as "Tunneling over MPLS LSPs").
>
> Cisco 6PE implements one specific flavor of "bgp-tunnel" (eg. "MP-BGP
> over IPv4" using "Tunneling over MPLS LSPs"). It is compliant to the
> specification of that flavor in "bgp-tunnel".

At this point, I think there is only requirement on "6PE", i.e., MPLS 
LSPs.  Other stuff could be specified in an extension document later 
on if needed.

> Your slides said:
> "some think bgp-tunnel has some cruft in it and would need a cleanup"
> 
> Because it specifies something for which there are multiple commercial
> implementations and multiple deployments in the field, the authors of
> bgp-tunnel have been trying very hard to get a WG home for this document
> in order to get feed-back on how to clean it up and finalise it. But the
> IETF decision was to hold onto that until the isp-scenario-analysis work
> is complete and identifies bgp-tunnel as a useful mechanism. 
> I understand that we will be able to discuss/progress bgp-tunnel in IETF
> (v6ops?) as soon as the isp-scenario-analysis document is completed. Do
> I understand the situation correctly?

Yes, I think this is correct.  Unless there are objections raised
about WG rough consnsus call, we're going forward with this.  The
question is just where (formally) -- here, in a new WG or a routing
area WG.  This is currently being considered, but if a routing area WG
would be willing to accept this, it would probably be best.

I think we (the IETF in general, or v6ops in particilar) want to ship 
off the document to the IESG before the next IETF.
 
> In the meantime, I'd like to invite folks who have identified cruft in
> bgp-tunnel to send comments on the list so we can start the potential
> cleanup process.

Agreed --the first step would probably need to be to remove the other
mechanisms than which "6PE" has, because we're only interested in that
specific scenario (and possibly the case where the MPLS control plane
has been upgraded to support IPv6 LSPs -- I don't know how this
relates to BGPTUNNEL spec, though.

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings