Pool member preference based on header

I think I accidentally cut that part out while removing superfluous log file entries: Client has a Node Preference of 10.1.150.171:28080. We parsed as 10.1.150.171:28080. Nodeport is 28080 Client wanted 10.1.150.171:28080. We were already load balanced to 10.1.150.186:28080. Do something about it. Client wanted 10.1.150.171:28080. We were already load balanced to 10.1.150.186:28080. We sent client to 10.1.150.171 28080 Client has a Node Preference of 10.1.150.171:28080. We parsed as 10.1.150.171:28080. Nodeport is 28080 TCL error: - invalid command name "28080" while executing "$nodeport == $Pool_Port" The pool is set up with 12 members. The load balancing method is "least connections (member)." There's no ratio or priority group set up. The health check is a simple tcp health check.