I have an Always-On architecture with two nodes and two AG. One of AG working correctly but other AG's listener only directs to the primary node. When I fail it to the secondary node, the application goes down. When I check everything looks fine.

select * from sys.dm_tcp_listener_states
select * from sys.availability_group_listeners
select * from sys.availability_group_listener_ip_addresses

enter image description here

  • 1
    Please add the listener definition (output of the DMVs you listed is fine) and type (network name, distributed network name), the environment you're running in (on-prem, AWS, Azure, etc), the output of netstat -ano -p tcp from the new primary replica where it isn't working. Dec 6, 2023 at 11:16
  • If AG is working correct, could also be the application not having a retry mechanism. Failovers cut current connections and the application has to reconnect. Dec 11, 2023 at 9:28


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.