We have updated the post and added a section to match this use case. You can find it under the subsection "Protecting TCP (non-HTTP) Services"