0

I'm not entirely sure whether this is even possible, but I thought I'd ask all the same.

So we have a service that connects to a streaming API which keeps a connection open and we receive JSON objects which we then throw into a message broker. Unfortunately, we can only have one connection open, which currently means we only have one VM running the instance.

Ideally, we'd like to have the ability to scale the service as and when so I guess I'm looking for some form of reverse load balancer?

Not sure if such a thing exists but any thoughts would be greatly appreciated

justcompile
  • 141
  • 3
  • Many client-side "load balancing" implementations rely on round-robin DNS, where multiple IP addresses are given for a DNS name, and the client chooses how to pick/choose the "next" IP address in the list. – Castaglia Apr 12 '16 at 17:40

1 Answers1

0

On the server-end you'd put something like a HAProxy in front of your app servers to load-balance across your servers.

On the client-end you'd create a round-robin DNS entry that points to your HAProxies, preferably with an automated healthcheck that takes one out of DNS if it fails.

Jason Martin
  • 5,023
  • 17
  • 24