You could run vpnv4 (4VPE) and vpnv6 (6VPE), and use: mpls label mode all-vrfs protocol all-afs per-vrf
Works for us. You will take a pps hit due to the aggregate triggered IP lookup. I too dream of and IPv6 control-plane. If that be SR, bring it on... Tim:> On Thu, Jan 2, 2014 at 3:29 AM, Steve Glendinning <st...@netthatworks.com>wrote: > Hi all, > > I've enabled 6PE on some of our rsp720-3cxl's (running SRE8) that carry > full global routes, and it appears to have allocated a separate MPLS label > for each IPv6 prefix in BGP (approx 16,000 of them). Consequently this is > chewing up significantly more FIB space, as each (global) IPv6 prefix now > uses 1 144-bit slot and 1 72-bit slot (so 3 of the 1M entries available). > > Is this right? It seems very wasteful, given that in the IPv4 world I only > have labels for my IGP routes and everything else is happy being sent > towards their favourite default route with just a "loopback" label on it. > > Some of the Cisco documentation mentions an "aggregate" IPv6 label, which > would make more sense. Is this available on 7600/6500? > > Thanks, > -- > Steve > _______________________________________________ > cisco-nsp mailing list cisco-nsp@puck.nether.net > https://puck.nether.net/mailman/listinfo/cisco-nsp > archive at http://puck.nether.net/pipermail/cisco-nsp/ > -- Tim:> _______________________________________________ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/