6PE is achieved with the following steps.
1. Exchange IPv6 reach ability information among 6PE routers with MP-BGP:
The 6PE routers exchange IPv6 prefixes over IPv4 MP-BGP sessions with a next-hop field that will be a IPv4 mapped IPv6 address. As well, the 6PEwill allocate a label to the IPv6 prefixes it connects to - theseassociated labels are used to forward traffic.
2. Transport IPv6 packets from the ingress 6PE router to the egress 6PE router over IPv4-signaled LSPs:
The ingress 6PE router forwards IPv6 traffic over a IPv4 LSP towards the egress 6PE router that resolves the IPv4 mapped IPv6 next-hop for the IPv6 prefix destination.
Here is a sample topology and the relevant configuration that enables 6PE. For simplicity, only the 6PE part of the configuration is provided and it is assumed that the MPLS/IGP infrastructure is already configured and operational.
PE1's config snippet:
neighbor 188.8.131.52 next-hop-self
neighbor 184.108.40.206 update-source 220.127.116.11
neighbor 2001::2 remote-as 2001
address-family ipv6 unicast
neighbor 18.104.22.168 activate
neighbor 22.214.171.124 send-label
neighbor 2001::2 activate
PE2's config snippet:
neighbor 126.96.36.199 remote-as 65512
neighbor 188.8.131.52 update-source 184.108.40.206
neighbor 220.127.116.11 remote-as 65512
neighbor 3001::2 remote-as 3001
address-family ipv6 unicast
neighbor 18.104.22.168 activate
neighbor 22.214.171.124 send-label
neighbor 3001::2 activate
The 6PE specific commands are marked in red. The 6PE BGP neighborship differs from a standard IPv6 neighborship by:
- Sending MPLSlabels for each prefix advertised
- The next-hop for the advertised IPv6 prefixes will be the IPv6 mapped addresses i.e. the advertised next-hop by PE1 would be ::ffff:126.96.36.199 (where 188.8.131.52 is its IPV4iBGPpeering address).
IPv6 connectivity must exist between the PEand its CE. Any IPv6 supported routing protocol can run between them to exchange route information. If it's other than IPv6 BGP, then the PEmust explicitly redistribute those routes into BGP so as to advertise to its remote peers.
With 6PE,some new commands are introduced and the output of a few existing commands are modified to display 6PE specific information.
The 6PE route updates carry a different SAFI(AFI=2, SAFI=4) which represents the capability to send labels - the outputs are modified to display this capability on the BGP neighbor session.
PE1#show ip bgp neigh 184.108.40.206
1 IP Address: 220.127.116.11, AS: 65512 (IBGP), RouterID: 18.104.22.168, VRF: default-vrf
Neighbor NLRI Negotiation:
Peer Negotiated IPV4 unicast capability
Peer Negotiated IPV6 unicast capability
Peer configured for IPV4 unicast Routes
Peer configured for IPV6 unicast Routes
Neighbor ipv6 MPLS Label Capability Negotiation:
Peer Negotiated ipv6 MPLS Label capability
Peer configured for ipv6 MPLS Label capability
The 'show ipv6 route' is added with a 'tags' option, which displays the labels associated with prefixes.
The sample output below is taken from PE2 - 2001:2::/64 is advertised by PE1, which is reachable via MPLS LSP 'v6' and the associated BGP next-hop is the IPv4 mapped address ::ffff:22.214.171.124
PE2#show ipv6 route tags
IPv6 Routing Table - 5 entries:
Type Codes - B:BGP C:Connected I:ISIS L:Local O:OSPF R:RIP S:Static
BGP Codes - i:iBGP e:eBGP
ISIS Codes - L1:Level-1 L2:Level-2
OSPF Codes - i:Inter Area 1:External Type 1 2:External Type 2
STATIC Codes - d:DHCPv6
Type IPv6 Prefix Next Hop Router Interface Dis/Metric Uptime src-vrf
1 Bi 2001:2::/64 :: lsp v6 200/1 15m37s -
label information: 794624(OUT)
No new debug keywords are introduced with 6PE. The existing 'debug ipv6 bgp update' will display MP-BGP updates with the associated label value.
PE-2#show ipv6 bgp 2001:2::/64
Number of BGP Routes matching display condition : 1
Status codes: s suppressed, d damped, h history, * valid, > best, i internal
Origin codes: i - IGP, e - EGP, ? - incomplete
Network Next Hop MED LocPrf Weight Path
*>i 2001:2::/64 ::ffff:126.96.36.199 1 100 0 2001 i
Last update to IP routing table: 0h29m44s
Route is advertised to 1 peers:
Note: 6PE functions in the default-vrf context i.e. the IPv6 CE peering cannot terminate in a VRF. 6VPE will support the requirement of the VRF connectivity.