cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
702
Views
0
Helpful
2
Replies

ibgp

Hello everyone,

How do i overcome the ibgp rule not to advertise a BGP advertisement learnt from one iBGP neighbor to another iBGP neighbor?

as in my topology i need router loopbacks advertised via ibgp.

all routers have ospf enabled and conectivity

all routers have ibgp peering

r2 and r3 cant see each other loopbacks

for every router config is as follow

 

r1:

neigh 10.1.12.2 remote-as 65000

neigh 10.1.14.4 remote-as 65000

neigh 10.1.13.3 remote-as 65000

network 1.1.1.1 mask 255.255.255.255

etc...

 

all the config is correct but i need to evercome the advertisment rule.

Thank you

 

1 Accepted Solution

Accepted Solutions

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello ,

you can use Route Reflector Servers or confederation AS with mini AS.

The first method is easier to implement.

 

On a selected router that should act as a BGP reflector  server for  some other iBGP peers you just need to add for each "client" peer the line

 

neighbor   x.x.x.x    route-reflector-client

 

This eliminates the need for a full mesh of iBGP sessions between clients of the RRS.

The RRS will propagate a client's prefix to all of its clients and ot all non clients iBGP peers.

The RRS will propagate to all clients a route learned from a non client iBGP peer.

 

The RRS devices need a full mesh of standard iBGP sessions between them.

 

For redundancy purposes each set of clients can be served by two RRS that can configure the same cluster-id  or a different one .

 

RRS adds two BGP attributes that allow fo safe reflection:

Originator -ID : BGP RID of the client that injected the prefix in the iBGP domain

Cluster list: it is a list of all the BGP RRS that the route has been propagated through when not set with cluster-id command the cluster-id = BGP RID of RRS.

in MP BGP implementations the neighbor ..... route-reflector-client command can be given in a specific address family and this can lead to RRS specialization in MPLS SP networks ( some RRS are used for global routing table others for VPNv4 services )

 

Edit:

advertising the local loopback in iBGP is not needed and not recommended advertise it in the underlying IGP OSPF

Hope to help

Giuseppe

View solution in original post

2 Replies 2

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello ,

you can use Route Reflector Servers or confederation AS with mini AS.

The first method is easier to implement.

 

On a selected router that should act as a BGP reflector  server for  some other iBGP peers you just need to add for each "client" peer the line

 

neighbor   x.x.x.x    route-reflector-client

 

This eliminates the need for a full mesh of iBGP sessions between clients of the RRS.

The RRS will propagate a client's prefix to all of its clients and ot all non clients iBGP peers.

The RRS will propagate to all clients a route learned from a non client iBGP peer.

 

The RRS devices need a full mesh of standard iBGP sessions between them.

 

For redundancy purposes each set of clients can be served by two RRS that can configure the same cluster-id  or a different one .

 

RRS adds two BGP attributes that allow fo safe reflection:

Originator -ID : BGP RID of the client that injected the prefix in the iBGP domain

Cluster list: it is a list of all the BGP RRS that the route has been propagated through when not set with cluster-id command the cluster-id = BGP RID of RRS.

in MP BGP implementations the neighbor ..... route-reflector-client command can be given in a specific address family and this can lead to RRS specialization in MPLS SP networks ( some RRS are used for global routing table others for VPNv4 services )

 

Edit:

advertising the local loopback in iBGP is not needed and not recommended advertise it in the underlying IGP OSPF

Hope to help

Giuseppe

thank you Giuseppe. just sorted with RR. 

Review Cisco Networking for a $25 gift card