Creates a new neighbor and makes it part of the peer group.
| remoteaddr | Specifies the IPv4 or IPv6 address of a BGP neighbor. |
| peer-group-name | Specifies a peer group. |
| multi-hop | Specifies to allow connections to EBGP peers that are not directly connected. |
N/A.
You can specify an IPv4 or IPv6 address for the BGP peer. The address can be a global unicast or a link-local address. IPv6 link-local remote addresses are supported only for EBGP single-hop peerings.
If you are adding an IPv4 peer to a peer group and no IPv4 address family capabilities are assigned to the specified peer group, the IPv4 unicast and multicast address families are automatically enabled for that peer group. If you adding an IPv6 peer to a peer group and no IPv6 address family capabilities are assigned to the peer group, you must explicitly enable the IPv6 address family capabilities you want to support.
Note
If the peer group or any member of the peer group has been configured with an IPv4 or IPv6 address family, the peer group only accepts peers that are configured to use that family. For example, if a peer group is configured for the IPv4 unicast address family, the switch will not allow you to add an IPv6 peer. LIkewise, an IPv6 peer group cannot accept an IPv4 peer.If the multihop keyword is not specified, the IP addresses of the EBGP speaker and peer must belong to the same subnet.
All the parameters of the neighbor are inherited from the peer group. The peer group should have the remote AS configured.
To add an existing neighbor to a peer group, use the following command:
configure bgp neighbor [all | remoteaddr] peer-group [peer-group-name | none] {acquire-all}If you do not specify acquire-all, only the mandatory parameters are inherited from the peer group. If you specify acquire-all, all of the parameters of the peer group are inherited. This command disables the neighbor before adding it to the peer group.
The following command creates a new neighbor and makes it part of the peer group outer:
create bgp neighbor 192.1.1.22 peer-group outer
The following example specifies how to create a neighbor peer group in a VRF (PE – CE neighbor session):
virtual-router <vr_vrf_name>
create bgp neighbor <remoteaddr> remote-AS-number <asNumber> {multi-hop}
create bgp neighbor <remoteaddr> peer-group <peer-group-name> {multi-hop}
delete bgp [{neighbor} <remoteaddr> | neighbor all ]
[create | delete] bgp peer-group <peer-group-name>
BGP maintains a separate RIB (RIB-In, RIB-Loc and RIB-Out) for each of the VRF it is configured to run. So routes received from a peer in VRF1 are not mixed up with routes from a peer in VRF2. Additionally, BGP routes in a VRF are regular IPv4 routes of address family ipv4. The BGP decision algorithm occurs inside a VRF and is not impacted by any BGP activity in other VRF.There can be two BGP neighbors with the same peer IP address in two different VRFs.
This command was first available in ExtremeXOS 10.1.
This command required a specific license in ExtremeXOS 11.1.
Support for IPv6 was added in ExtremeXOS 12.6 BGP.
Support for L3 VPN was added in ExtremeXOS 15.3.
This command is available on platforms that support the appropriate license. For complete information about software licensing, including how to obtain and upgrade your license and which licenses support the BGP feature,see the ExtremeXOS and Switch Engine 31.7 Feature License Requirements document.