Hi Arun,

It's for EVPN only, but using the same concept that has been developed for L3.

If we have both L2 and L3, it would be "ships in the night", and 
https://tools.ietf.org/html/draft-lin-bess-evpn-irb-mcast-02 is one way to 
optimize things. That draft talked about using MVPN when a bridge domain is not 
connected everywhere.

Jeffrey

> -----Original Message-----
> From: Arun Prakash (arprakas) [mailto:[email protected]]
> Sent: Friday, September 02, 2016 5:17 PM
> To: Jeffrey (Zhaohui) Zhang <[email protected]>; Swadesh Agrawal
> (swaagraw) <[email protected]>; Eric Rosen <[email protected]>; Martin
> Vigoureux <[email protected]>; [email protected]
> Cc: [email protected]; Nikhil Shetty
> (nikshett) <[email protected]>
> Subject: Re: [bess] Poll for adoption: draft-zzhang-bess-evpn-bum-
> procedure-updates-03
> 
> Hi Jeffrey,
> 
> Thank you for the explanation. Am not clear if this draft is only for L2
> or can it be used for L3 as well?
> What if we have both L2 and L3 multicast ?
> 
> Thanks
> Arun
> 
> On 8/28/16, 7:08 PM, "BESS on behalf of Jeffrey (Zhaohui) Zhang"
> <[email protected] on behalf of [email protected]> wrote:
> 
> >Hi Swadesh,
> >
> >"Enable MVPN safi and enhance it for EVPN" is not a trivial thing that
> >can be cleanly done on top of MVPN. You'd end up complicating MVPN
> >procedures a lot for EVPN specific things. Instead, we can apply the same
> >principles to EVPN safi and procedures, and that's what this draft is
> >about.
> >
> >BTW, we don't yet see the need of the equivalent of MVPN type-5 routes
> >for EVPN. Draft-sajassi-bess-evpn-igmp-mld-proxy already proposed SMET
> >route which is very similar to MVPN type-6/7 (to be exact, a combination
> >of type-4/6/7). That is explained in
> >draft-zzhang-bess-mvpn-evpn-cmcast-enhancements.
> >
> >Jeffrey
> >
> >> -----Original Message-----
> >> From: Swadesh Agrawal (swaagraw) [mailto:[email protected]]
> >> Sent: Friday, August 26, 2016 3:06 AM
> >> To: Eric Rosen <[email protected]>; Martin Vigoureux
> >> <[email protected]>; [email protected]
> >> Cc: [email protected]
> >> Subject: Re: [bess] Poll for adoption: draft-zzhang-bess-evpn-bum-
> >> procedure-updates-03
> >>
> >> Why we need to copy MVPN procedures in EVPN. Why not enable MVPN safi
> >>and
> >> enhance it for any EVPN specific requiremnt. Also going in this
> >>direction
> >> may require to add 5,6 and 7 routes as well in future.
> >>
> >> Regards
> >> Swadesh
> >>
> >> On 8/25/16, 2:22 PM, "Eric C Rosen" <[email protected]> wrote:
> >>
> >> >Support.
> >> >
> >> >MVPN has a number of multicast features that are valuable, but haven't
> >> >yet been incorporated into EVPN.  This draft takes a comprehensive
> look
> >> >at the MVPN  procedures that provide those features, and shows how to
> >> >adapt them to EVPN.   This seems quite valuable to me, and thus I
> think
> >> >the WG should adopt this draft.
> >> >
> >> >_______________________________________________
> >> >BESS mailing list
> >> >[email protected]
> >> >https://www.ietf.org/mailman/listinfo/bess
> >
> >_______________________________________________
> >BESS mailing list
> >[email protected]
> >https://www.ietf.org/mailman/listinfo/bess

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

Reply via email to