Hi Friends,
I am facing a weird issue with BGP routing in Nexus 3k (NX OS 6).
Router A (AS 6500) and B (AS 6600) are eBGP peers.
Router A:
int Eth1/1.100 (no vrf) IP 1.1.1.1 & peer 2.2.2.2
int Eth1/1.200 (vrf test) IP 10.10.10.10 & peer 20.20.20.20
Router A BGP config:
router bgp 64100
log-neighbor-changes
address-family ipv4 unicast
neighbor 2.2.2.2 remote-as 6600
local-as 65500 no-prepend replace-as
timers 1 3
address-family ipv4 unicast
soft-reconfiguration inbound always
vrf test
address-family ipv4 unicast
neighbor 20.20.20.20 remote-as 6600
local-as 65500 no-prepend replace-as
address-family ipv4 unicast
soft-reconfiguration inbound always
There are 2 instances of BGP running on Router A with same local & remote AS; one in global routing table and other in VRF test. The BGP adjacency comes up fine between the peers one at a time; however I have noticed that only 1 BGP session remains in "established" state at a given point of time - either the one in global routing table or the one in VRF Test. And the other one goes to Idle / Active state. If I clear BGP session for the idle/active one, it comes up and the other one goes down.
Is this normal? Or is there flaw in the design? How do I overcome this scenario?
