TCP connection health checks always successful

I’m trying to configure a TCP health probe to a TCP service in istio.

Issue I’m facing however is that even when the service is fully down (replicas = 0) the TCP handshake will still be successful making the basic TCP health probe think the service is up.

I tried looking across the documentation to change this behavior but there doesn’t seem to be a way to make istio report a TCP connection error when the service isn’t available.

Did I miss anything?