On 1/5/2016 10:50 AM, Vivek Venkatraman wrote:
>
> On Tue, Jan 5, 2016 at 5:44 AM, Lou Berger <[email protected]
> <mailto:[email protected]>> wrote:
>
>     Vivek,
>
>     > I think there will be quite a few things to discuss and agree upon
>     > when the import/export functionality (i.e., PE behavior) is built on
>     > top of these changes (and the VRF support that is being introduced
>     > into BGP),
>
>     I certainly agree that there's a lot of discussion to be had WRT
>     import/export functionality, and this is part of the reason why this
>     core set doesn't include any of those changes.  The intended scope of
>     this patch set was limited to fixing and completing what was there in
>     bgp_mplsvpn got the VPN_SAFI and  adding the equivalent function for
>     ENCAP_SAFI.  So I think the PE functionality related discussions
>     should
>     not gate acceptance of these fixes/parallel additions.
>
>
> Yes, I am in agreement. The PE functionality needs these patches as
> well as the BGP VRF work that Vipin is currently working on (which is
> based on the base VRF functionality).  They will also need some MPLS
> (or labeled route) integration if MPLS encap is used and we have some
> code in the works on that front.
>
Yes, linux data plane integration will be interesting.  To date all our
label work has used the NVO3 approach, i.e., control plane only.

> Those discussions can certainly happen in parallel and need not gate
> acceptance of these patches, IMO.
> Note: I haven't looked at your UI (vty) changes yet and as you mention
> below, those changes may elicit opinion/discussion from many.
>
>

Yup.  Would be nice to standardize on a single afi/safi format, but then
we need to consider the compatibility implications...

Lou



_______________________________________________
Quagga-dev mailing list
[email protected]
https://lists.quagga.net/mailman/listinfo/quagga-dev

Reply via email to