iBGP configuration.
Generally, internal BGP (IBGP) speakers need to be fully meshed because IBGP does not
readvertise updates to other IBGP speakers.
This full mesh requirement does not scale when there are large number of IBGP speakers,as within an AS n*(n-1)/2 unique ibgp sessions must be maintained which doesn’t scale with finite resources in both bandwidth & router resources.To solve this,BGP route reflectors and BGP confederations have been developed.
In this example,i will configure Delhi,Mumbai as route reflector servers and BLR,Chennai as route reflector clients.
On Device Delhi,Mumbai the RouteReflectors, you must form peer relationships with all of the IBGP speakers by including the neighbor statement for the clients (Devices BLR and Chennai) and the nonclients (if any). You must also include the cluster statement and a cluster identifier.This example uses the loopback interface IP address of the RR as Cluster ID.
On Devices BLR and Chennai, the RR clients, you only need one neighbor statement that forms a peer relationship with the RR’s Delhi,Mumbai.
Route Reflector:Step-by-Step Procedure
-configure AS Number
root@Mumbai# set routing-options autonomous-system 3333
-Configure BGP, including the cluster identifier and neighbor relationships with all IBGP speakers in the autonomous system (AS). Also apply the policy that redistributes OSPF routes into BGP.
root@Mumbai# set protocols bgp group internal_peers type internal
root@Mumbai# set protocols bgp group internal_peers local-address 2.2.2.2
root@Mumbai# set protocols bgp group internal_peers export send-ospf
root@Mumbai# set protocols bgp group internal_peers neighbor 1.1.1.1
root@Mumbai# set protocols bgp group internal_peers neighbor 3.3.3.3
root@Mumbai# set protocols bgp group internal_peers neighbor 4.4.4.4
-Configure the policy that redistributes OSPF routes into BGP.
root@Mumbai# set policy-options policy-statement send-ospf term 1 from protocol ospf
root@Mumbai# set policy-options policy-statement send-ospf term 1 then accept
Repeat the same for Delhi Router since we want it as RR as well.
Route Reflector Client:Step-by-Step Procedure
-configure AS Number
root@BLR# set routing-options autonomous-system 3333
-Configure the BGP neighbor relationship with the RR. Also apply the policy that
redistributes OSPF routes into BGP.
root@BLR# set protocols bgp group internal_peers type internal
root@BLR# set protocols bgp group internal_peers local-address 4.4.4.4
root@BLR# set protocols bgp group internal_peers export send-ospf
root@BLR# set protocols bgp group internal_peers neighbor 1.1.1.1
root@BLR# set protocols bgp group internal_peers neighbor 2.2.2.2
-Configure the policy that redistributes OSPF routes into BGP.
root@BLR# set policy-options policy-statement send-ospf term 1 from protocol ospf
root@BLR# set policy-options policy-statement send-ospf term 1 then accept
Repeat the same for chennai Router since we want it as RR client as well,but give neighbor ID as mumbai (2.2.2.2)
Verification:
• Verifying BGP Neighbors.
sh bgp neighbor (i didnt get rr client in output,why ?)
• Verifying BGP Groups.
sh bgp group
• Verifying BGP Summary Information.
sh bgp summary
• Verifying Routing Table Information.
sh route (to check we learn any routes frm protocol bgp)
Network Automation Blog 