When BGP receives a same prefix from multiple neighbors and if the next hop is also of the same subnet, then the route keeps flapping and user will see a continuous addition and deletion of the prefix.

Products

• VMware NSX

Issue/Introduction

Symptoms:

• Traffic drop for prefixes that are getting continuously added/deleted
• In Standby Edge syslog logs, many FIB route updates can be seen:
  • 2024-06-25T16:07:55.723Z edge NSX 1 ROUTING [nsx@6876 comp="nsx-edge" subcomp="nsxa" s2comp="routing" level="INFO"] FIB update 2 update 10.22.42.0/23 add
    2024-06-25T16:07:55.723Z edge NSX 1 ROUTING [nsx@6876 comp="nsx-edge" subcomp="nsxa" s2comp="routing" level="INFO"] FIB update 2 update 10.122.52.0/20 add
    2024-06-25T16:07:55.723Z edge NSX 1 ROUTING [nsx@6876 comp="nsx-edge" subcomp="nsxa" s2comp="routing" level="INFO"] FIB update 2 update 10.192.10.0/20 add
    2024-06-25T16:07:55.723Z edge NSX 1 ROUTING [nsx@6876 comp="nsx-edge" subcomp="nsxa" s2comp="routing" level="INFO"] FIB update 2 update 10.54.122.0/23 add
    
    
• FRR logs are filled with similar logs:
  • 2024/06/24 04:01:54.353972 ZEBRA: 0:192.168.20.0/31 has Nexthop(192.168.20.1/32) Type: Nexthop depending on it, evaluating 2390141104:2390141084
2024/06/24 04:01:54.353989 ZEBRA: zebra_rnh_resolve_nexthop_entry: 0:192.168.20.1/32 Possible Match to 192.168.20.0/31
2024/06/24 04:01:54.353995 ZEBRA: zebra_rnh_remove_from_routing_table: 0:192.168.20.1/32 removed from tracking on 192.168.20.0/31
2024/06/24 04:01:54.354002 ZEBRA: zebra_rnh_store_in_routing_table: 0:192.168.20.1/32 added for tracking on 

 

• BGP or Static prefixes and the next-hop are in the same subnet, examples are below :
  • S 192.168.12.0/32 [3/0] via 192.168.12.1
  • b  > * 172.16.190.4/32 [20/0] via 172.16.190.6

Environment

VMware NSX-T Data Center 3.x

Resolution

Workaround:

Add an inbound route-map that filters the BGP prefix that is in the same subnet as the static route next-hop.

 

This issue is resolved in NSX-T 4.x or higher.