Support Center > Search Results > SecureKnowledge Details
Adding/changing an IP address on a VRRP interface causes the BGP connections to restart on that VRRP interface
Symptoms
  • Adding/changing an IP address on a VRRP interface (e.g., IP address is added to a new VLAN that is configured on a VRRP interface) causes the BGP connections to restart on that VRRP interface.

    Example:

    1. BGP neighborship is established between VRRP cluster and a BGP peer
    2. BGP routes are successfully re-distributed
    3. Output of Clish command "show bgp peers" shows "State" as "Established"
    4. Added a VLAN interface on interface a VRRP interface (e.g., eth3)
    5. Output of Clish command "show bgp peers" shows "State" as "Active"
  • BGP traces during the issue showed:
    bgp_ifaddr_change: peer <IP_of_BGP_Peer> (External AS <N>) restarted, shared interface changed

Cause

When an IP address is added/changed on an interface (e.g., IP address is added to a new VLAN), RouteD daemon receives an interface event from the OS.
This triggers RouteD daemon to see whether the reported change affects any enabled routing protocol, in this case - BGP.
BGP checks if it is affected by IP address change on an interface by going over known BGP peers and seeing if the interface IP address differs from the local address for that BGP peer. If the address has changed, then BGP peer connection should be restarted.
In this case, the local address for that peer is the VRRP VIP, but it was compared to the physical IP address of the interface - since they are different, the BGP connection was restarted.


Solution
Note: To view this solution you need to Sign In .