Routing Bits

Several BGP timers can be used to tweak the BGP convergence in an IP/MPLS/VPN Network. A good understanding and side by side comparison will allow you to tweak BGP behavior.

BGP Scan Time:
#bgp scan-time {scanner-interval}

• Configures  import  processing  of  IPV4  unicast  routing  information  from  BGP routes  into  routing  tables.
• The BGP scanner walks the BGP table and confirms the reachability of next hops.
• The BGP scanner process is also responsible for conditional advertisement check and performing route dampening.
• The Max / Default value = 60 sec.  Range = 5-60 sec.
• Lowering the scan-time, can greatly improve BGP convergence, to almost IGP-like behaviour.

.

BGP Timers:
#timers bgp {keepalive} {holdtime}
#neighbor {IP|Peer-Group} timers {keepalive} {holdtime}

• Keepalive    - Frequency,  in  seconds,  with  which  the  Cisco  IOS  software  sends  keepalive messages to its peer.
• The default is 60 seconds.
• Holdtime    - Interval,  in  seconds,  after  not  receiving  a  keepalive  message  that  the  software declares a peer dead.
• The default is 180 seconds.
• Only the holdtime value is communicated in the BGP Open messages.
• Smallest configured holdtime value between BGP peers is used by both peers and used to determine the keepalive.
• Can be set on process-level or per neighbor. Neighbor-level command overrides process-level.

.

BGP Update Interval:
#neighbor {IP|peer-group} advertisement-interval {sec}

• To set the interval between the sending of BGP routing updates.
• Lowered value  can improve convergence,at the cost of consuming considerable resources.
• Not suggested to use value to low in jittery networks.
• Range 0 to 600 seconds
• The default (5 sec or iBGP peers) (30 sec for eBGP peers)

.

BGP Fast External Fallover:
#no bgp fast-external-fallover

• BGP fallover for external peers is triggered by a session flap.
• By default the timming of the fallover is based upon the receipt of an interface change notification (ICN) when the flap takes place.  It might not always desirable to trigger a fallover every time the interface state changes, ie a flapping link.
• Disabling BGP fast external fallover changes the trigger from the ICN to the value of   the BGP hold-timer.
• Command is enabled by default.

.

BGP Fast Peering Session:
ip prefix-list FILTER seq 5 permit 192.168.0/16 le 32
!
route-map CHECK-NBR permit 10
match ip address prefix-list FILTER
!
router bgp 45000
neighbor 192.168.1.2 remote-as 40000
neighbor 192.168.1.2 fall-over route-map CHECK-NBR

• Used to enable BGP to monitor the peering session of a specified neighbor for adjacency changes and to deactivate the peering session.
• BGP fast peering session deactivation is event driven and is configured on a per-neighbor basis.
• Adjacency changes are detected, and terminated peering sessions are deactivated in between the BGP scanning intervals.
• A route-map can be used to deactivate the peering session based on a specific prefix.
• Only the “match ip address” and “match source-protocol” commands are supported in the fall-over route map.

.

BGP Next-Hop Address Tracking:
#no bgp nexthop trigger enable

• BGP prefixes are automatically tracked as peering sessions are established.
• Next-hop changes are rapidly reported to the BGP routing process as they are updated in the Routing Information Base (RIB).
• This optimization improves overall BGP convergence by reducing the response time to next-hop changes for routes installed in the RIB.
• When a best-path calculation is run in between BGP scanner cycles, only next-hop changes are tracked and processed.
• Command is enabled by default.