NimbostratusWe have deployed LTM(11.6.0 HF5) in inline mode. When I try accessing direct node on a specific port to test the application, it's working fine, But when I am try accessing the application via VIP, some Routing Error is there.
3-way handshake is happening with Virtual server but LB is sending Rst-ACK immediately. Any Suggestions?
Below is the screenshot: -
EmployeeRouting on the client side obviously looks fine. Do you see any server side traffic?
NimbostratusThere is no server side connections for the above pasye wireshark.. But yes from LB I am able to reach servers on required ports. Also when I bypass LB and hit direct server by using forwarding VS.. Connection seems fine....
EmployeeI guess more specifically I'm asking if you see any traffic on the server traffic, even just an attempt (initial SYN)?
Also when you say
Also when I bypass LB and hit direct server by using forwarding VS
Are you saying that you're bypassing the load balancer and sending traffic through a forwarding virtual server? Can you elaborate on that?
NimbostratusNo, there is no initial syn for server side connection..
By bypassing the LB i mean that we have deployed LTM in inline mode.. LB selfip is the default gayeway for servers.. traffic is just being forwarded via LTM(without any load balancing mechanism) with the help if ip forwarding virtual servers..
NimbostratusI hope it clears now..
EmployeeOkay, so we now know that this is internal traffic going out. Correct me if I'm wrong here.
We know that if you configure a wildcard (0.0.0.0:0/0) IP forwarding virtual server and create a default route on the BIG-IP that points to your outbound router, that traffic passes through that correctly.
But if you configure a standard virtual server (presumably also 0.0.0.0:0/0) and assign it a pool (presumably the same outbound router), that this doesn't work.
Am I close?
NimbostratusIssue was with configuration, Route domain ID was missing in the Nodes. :(