> Date: Sun, 4 Jan 2009 20:09:08 +0100 > From: Daniel Roesen <d...@cluenet.de> > Sender: juniper-nsp-boun...@puck.nether.net > > On Fri, Jan 02, 2009 at 09:03:05PM +0800, Mark Tinka wrote: > > local-address 2001:db8::3:ff7b; > > family inet6-vpn { > > apply-groups BGP-OUTBOUND-POLICY6-L3VPN6; > > unicast; > > } > > export BGP-OUTBOUND-POLICY6; > > peer-as 65000; > > neighbor 2001:db8::5:ff79; > > neighbor 2001:db8::5:ff78; > > This would "native" IPv6 MPLS L3VPN support, which isn't there (no > customer demand, yaddayadda). There is only 6PVE support (6PE hack for > L3VPNs), which means you transport the IPv6 L3VPN NRLI within IPv4 > BGP transport sessions, not IPv6 sessions (and thus having IPv4 BGP > NEXT_HOPs).
The number of features missing for IPv6 from all vendors) is very frustrating and several are far more serious than this one. (Actually it's not an issue for us.) Unless your iBGP mesh is very small, it's far easier to run a single mesh. We have always exchanged our IPv6 routes over an IPv4 iBGP mesh just to make things more manageable. When we see really complete IPv6 support in JunOS, we will move both v4 and v6 announcements to run over a single IPv6 mesh, but I suspect that will not be something for us to worry about for at least a couple more years and probably until after I retire. -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: ober...@es.net Phone: +1 510 486-8634 Key fingerprint:059B 2DDF 031C 9BA3 14A4 EADA 927D EBB3 987B 3751
pgpK4EyEeK6Ow.pgp
Description: PGP signature
_______________________________________________ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp