I am currently looking into migrating my current system over to Google Cloud. We have multiple on field devices that contact a specific server and log data. each of these servers is running custom software that processes the data.

The custom software and the field devices communicate to each other over TCP. There a few different versions and types running so we need to point the field devices to the correct servers manually.

I would like to set up multiple Google Cloud Engine VMs in the same region/zone under a single IP address and address them via port forwarding. However, It does not look like google supports this. I can do what I need to do using static IPs but there are limits to IP address use and we want to be able to scale correctly with more VMs as we gain more clients.

What is the best way to talk between the clients and servers while keeping IP use down?

1 Answer 1


If your devices and servers speak HTTP, you can use an HTTP(S) load balancer. Otherwise you can use a TCP Proxy load balancer, but the range of allowed port numbers is restricted.

  • They communicate over TCP. I had a look at the load balancers and didn't really see one to do what we needed. Does the TCP Proxy load balancer allow for manual server selection? We don't need auto load balancing and that seems to be what this does.
    – alex.brown
    Aug 19, 2019 at 7:21
  • @alex.brown I wonder if you did look at it? You certainly can select your instances manually; this was shown in the example! Aug 19, 2019 at 7:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.