Users Guide

Neighbor soft-reconfiguration
BGP soft-reconfiguration allows for fast route changes. Changing routing policies requires a reset of BGP sessions or the TCP
connection, for the policies to take effect.
Resets cause undue interruption to traffic due to the hard reset of the BGP cache, and the time it takes to reestablish the
session. BGP soft-reconfiguration allows for policies to apply to a session without clearing the BGP session. You can perform a
soft-reconfiguration on a per-neighbor basis, either inbound or outbound. BGP soft-reconfiguration clears the policies without
resetting the TCP connection. After configuring soft-reconfiguration, use the clear ip bgp command to make the neighbor
use soft reconfiguration.
When you enable soft-reconfiguration for a neighbor and you run the clear ip bgp soft in command, the update
database that is stored in the router replays and updates are reevaluated. With this command, the replay and update process
trigger only if a route-refresh request is not negotiated with the peer. If the request is negotiated after using the clear ip
bgp soft in command, BGP sends a route-refresh request to the neighbor and receives all the updates of the peer.
To use soft reconfiguration, or soft reset without preconfiguration, both BGP peers must support soft route refresh. The soft
route refresh advertises in the OPEN message sent when the peers establish a TCP session. To determine whether a BGP
router supports this capability, use the show ip bgp neighbors command. If a router supports the route refresh capability,
the Received route refresh capability from peer message displays.
1. Enable soft-reconfiguration for the BGP neighbor and BGP template in ROUTER-BGP mode. BGP stores all the updates
that the neighbor receives but does not reset the peer-session. Using this command starts the storage of updates, which is
required for inbound soft reconfiguration.
neighbor {ip-address} soft-reconfiguration inbound
2. Enter Address Family mode in ROUTER-NEIGHBOR mode.
address-family {[ipv4 | ipv6] [unicast]}
3. Configure soft-configuration for the neighbors belonging to the template.
soft-reconfiguration inbound
4. Clear all information or only specific details in EXEC mode.
clear ip bgp {neighbor-address | *} [soft in]
● * — Clears all peers.
● neighbor-address— Clears the neighbor with this IP address.
Soft-reconfiguration of IPv4 neighbor
OS10(conf-router-neighbor)# address-family ipv4 unicast
OS10(conf-router-bgp-neighbor-af)# soft-reconfiguration inbound
OS10(conf-router-bgp-neighbor-af)# end
OS10# clear ip bgp 10.2.1.2
Soft-reconfiguration of IPv6 neighbor
OS10(conf-router-neighbor)# address-family ipv6 unicast
OS10(conf-router-bgp-neighbor-af)# soft-reconfiguration inbound
OS10(conf-router-bgp-neighbor-af)# end
OS10# clear ip bgp 2001:0000:3221:DFE8:63::FEAB
Redistribute iBGP route to OSPF
When you configure the system redistribute BGP routes to OSPF, by default, the system redistributes only the eBGP routes.
To redistribute iBGP routes to OSPF, configure a route-map and apply it to the redistribute command under the OSPF
configuration.
Sample configuration
The following sample topology has two switches, Core 1 and Core 2, that are connected to each other and share routes using
OSPF. A border router BR is connected to Core 1 and shares routes using internal BGP. Core 1 redistributes the routes that are
learned by iBGP to OSPF to Core 2.
600
Layer 3