Foundry Configuration Guide for the FESX, FSX, and FWSX
21 - 34 © Foundry Networks, Inc. December 2005
the CLUSTER_LIST attribute to the route. Other route reflectors who receive the route from an IBGP
neighbor add their cluster IDs to the front of the route’s CLUSTER_LIST. If the route reflector does not have a
cluster ID configured, the Layer 3 Switch adds its router ID to the front of the CLUSTER_LIST.
• If Layer 3 Switch configured as a route reflector receives a route whose CLUSTER_LIST contains the route
reflector’s own cluster ID, the route reflector discards the route and does not forward it.
Configuration Procedures
To configure a Foundry Layer 3 Switch to be a BGP4 route reflector, use either of the following methods.
NOTE: All configuration for route reflection takes place on the route reflectors, not on the clients.
Enter the following commands to configure a Foundry Layer 3 Switch as route reflector 1 in Figure 21.2 on
page 21-33. To configure route reflector 2, enter the same commands on the Layer 3 Switch that will be route
reflector 2. The clients require no configuration for route reflection.
FESX424 Router(config-bgp-router)# cluster-id 1
FESX424 Router(config-bgp-router)# neighbor 10.0.1.0 route-reflector-client
FESX424 Router(config-bgp-router)# neighbor 10.0.2.0 route-reflector-client
Syntax: [no] cluster-id <num> | <ip-addr>
The <num> | <ip-addr> parameter specifies the cluster ID and can be a number from 1 – 4294967295 or an IP
address. The default is the router ID. You can configure one cluster ID on the router. All route-reflector clients for
the router are members of the cluster.
NOTE: If the cluster contains more than one route reflector, you need to configure the same cluster ID on all the
route reflectors in the cluster. The cluster ID helps route reflectors avoid loops within the cluster.
To add an IBGP neighbor to the cluster, enter the following command:
Syntax: neighbor <ip-addr> route-reflector-client
For more information about the neighbor command, see “Adding BGP4 Neighbors” on page 21-12.
By default, the clients of a route reflector are not required to be fully meshed; the routes from a client are reflected
to other clients. However, if the clients are fully meshed, route reflection is not required between clients.
If you need to disable route reflection between clients, enter the following command. When the feature is disabled,
route reflection does not occur between clients but reflection does still occur between clients and non-clients.
FESX424 Router(config-bgp-router)# no client-to-client-reflection
Enter the following command to re-enable the feature:
FESX424 Router(config-bgp-router)# client-to-client-reflection
Syntax: [no] client-to-client-reflection
Configuring Confederations
A confederation is a BGP4 Autonomous System (AS) that has been subdivided into multiple, smaller ASs.
Subdividing an AS into smaller ASs simplifies administration and reduces BGP-related traffic, thus reducing the
complexity of the Interior Border Gateway Protocol (IBGP) mesh among the BGP routers in the AS.
The Foundry implementation of this feature is based on RFC 3065.
Normally, all BGP routers within an AS must be fully meshed, so that each BGP router has interfaces to all the
other BGP routers within the AS. This is feasible in smaller ASs but becomes unmanageable in ASs containing
many BGP routers.
When you configure BGP routers into a confederation, all the routers within a sub-AS (a subdivision of the AS) use
IBGP and must be fully meshed. However, routers use EBGP to communicate between different sub-ASs.
To Next Page
Loading...
Need help?
General
