Extensive peering arrangements are a critical component to providing outstanding connectivity and performance for the customers as well as increased service resilience through diversified routing.
Peering is the bedrock of every network and all service providers actively expand their peering capacity in line with traffic growth to maintain world class performance.
BGP is an exterior gateway protocol (EGP), used to perform interdomain routing in TCP/IP networks. A BGP router needs to establish a connection (on TCP port 179) to each of it's BGP peers before BGP updates can be exchanged. The BGP session between two BGP peers is said to be an external BGP (eBGP) session if the BGP peers are in different autonomous systems (AS). A BGP session between two BGP peers is said to be an internal BGP (iBGP) session if the
BGP peers are in the same autonomous system.
By default, the peer relationship is established using the IP address of the interface closest to the peer router. However, using the neighbor update-source command, any operational interface, including the loopback interface, can be specified to be used for establishing TCP connections. This method of peering using a loopback interface is useful since it will not bring down the BGP session when there are multiple paths between the BGP peers, which would otherwise result in tearing down the BGP session if the physical interface used for establishing the session goes down. In addition to that, it also allows the routers running BGP with multiple links between them to load balance over the available paths.
The backbone cost burden associated with settlement-free peering traffic exchange should be equitably shared. Regardless of the direction or type of traffic exchanged between the networks, the routing practices and location or interconnection points should be such that each party bears a reasonably equal share of backbone costs.
In the case that the peering party has significant excess inbound traffic volume to deliver to (significantly in excess of that being sent outbound by to that party) then the parties should agree the routing behaviour expected of one another within reasonable technical, and operationally-feasible bounds.
When the peering partner has network presence in multiple countries or continents where also has Points Of Presence, it will be desirable to open peering legs in at least two locations, and maybe more, subject to the cost/benefits of PoP build-out. This is to ensure that the traffic is exchanged optimally in the respective countries/continents and peering interconnections benefit both partners equally.
Peers must be directly connected when using eBGP. If they are not directly connected, the neighbor ebgp-multihop command must be used and a path through an IGP or static route to reach the peer must exist in order for the routers to establish neighbor relationship.
Use the show ip bgp neighbors command to display information about the TCP and Border Gateway Protocol (BGP) connections and verify if the BGP peer is established. The output of the show ip bgp neighbors command below shows the BGP state as 'Established', which indicates that the BGP peer relationship has been established successfully.
R1-AGS# show ip bgp neighbors | include BGP
BGP neighbor is 10.10.10.2, remote AS 400, internal link
BGP version 4, remote router ID 18.104.22.168
BGP state = Established, up for 00:04:20
BGP table version 1, neighbor version 1
The show ip bgp neighbors command has been used above with the modifier | include BGP. This makes the output more readable by filtering the the command output and displaying the relevant parts only.
In addition, the show ip bgp summary command can also be used to display the status of all BGP connections, as shown below.
R1-AGS(9)# show ip bgp summary
BGP router identifier 10.1.1.2, local AS number 400
BGP table version is 1, main routing table version 1
Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd
10.10.10.2 4 400 3 3 1 0 0 00:00:26 0
You can also configure eBGP/iBGP using a loopback address (or any other operational interface). Loopback interfaces are used in this manner to guarantee reachability in networks with multiple paths as shown in Sample Configurations of the Load Sharing Using the Loopback Address as a BGP Neighbor section of Load Sharing with BGP in Single and Multihomed Environments.
This case shows how to achieve load sharing when multiple links exist between a remote AS and . These links are terminated in one router at the and on multiple routers at remote ASs in a single-homed BGP environment.