The switch extends the BGPv4 process to support the exchange of IPv6 routes using BGPv4 peering. BGP+ is an extension of BGPv4 for IPv6, which is indicated using the Address Family Identifier (AFI) in the BGP header.
The switch supports capabilities for AFI with the following values: 1 (IPv4) and 2 (IPv6). If the switch receives an OPEN message advertising an AFI with a different value, the connection is closed and a BGP notification message is sent to the peer mentioning unsupported capability.
BGP+ is only supported on the global VRF instance.
Note
Ensure you configure IPv6 forwarding for BGP+ to work.
Note that the BGP+ support on the switch is not an implementation of BGPv6. Native BGPv6 peering uses the IPv6 Transport layer (TCPv6 ) for establishing the BGPv6 peering, route exchanges, and data traffic.
The switch supports the exchange of IPv6 reachability information over IPv4 transport. To support BGP+, the switch supports two BGP protocol extensions, standards RFC 4760 (multi-protocol extensions to BGP) and RFC 2545 (MP-BGP for IPv6). These extensions allow BGPv4 peering to be enabled with IPv6 address family capabilities.
The implementation of BGP+ on the switch uses an existing TCPv4 stack to establish a BGPv4 connection. Optionally, nontransitive BGP properties are used to transfer IPv6 routes over the BGPv4 connection. Any BGP+ speaker has to maintain at least one IPv4 address to establish a BGPv4 connection.
Different from IPv4, IPv6 introduces scoped unicast addresses, identifying whether the address is global or link-local. When BGP+ is used to convey IPv6 reachability information for interdomain routing, it is sometimes necessary to announce a next hop attribute that consists of a global address and a link-local address. For BGP+, no distinction is made between global and site-local addresses.
The BGP+ implementation includes support for BGPv6 policies, including redistributing BGPv6 into OSPFv3, IS-IS, RIPng, and advertising OSPFv3, IS-IS, RIPng, IPv6 static and local routes into BGPv6 (through BGP+). It also supports the aggregation of global unicast IPv6 addresses.
When configuring BGP+ on the router that is enabled only for IPv6 (the router does not have an IPv4 address), then BGP router ID must be manually configured for the router.
BGP+ does not support confederations. You can configure confederations for IPv4 routes only.
The basic configuration of BGP+ is the same as BGPv4 with one additional parameter added and some existing commands altered to support IPv6 capabilities. You can enable and disable IPv6 route exchange by specifying the address family attribute as IPv6. Note that an IPv6 tunnel is required for the flow of IPv6 data traffic.
When you use BGP+ you must configure an IPv6 tunnel and static routes at BGP+ peers.
When BGP+ peers advertise route information, they use Update messages to advertise route information.
These RTM routes contain next-hop addresses from the BGP peer that the route was learned from.
The static routes correlate the next-hop addresses represented by the IPv4–mapped IPv6 address to a specific outgoing interface.
Following is one way to express a static route in an IPv6–configured tunnel for BGP+:
ipv6 route 2001:DB8:0:0:0:ffff:192.0.2.0/24 cost 1 tunnel 10 where 2001:DB8:0:0:0:ffff:192.0.2.0 is the IPv4-mapped IPv6 address of the BGP peer at 192.0.2.0