I've heard about the dynamic scaling with instance VM's, but as a large global cloud provider network engineer myself, I'm super interested in geeking out over CGI'S scalable cloud architecture strategy for the PU.

This game most likely has some particular requirements of its network compared to other applications, due to the fidelity and scope.

1. super low latency (gaming), which means intelligent compression, CDN delivery of common assets, telemetry, etc

2. high availability, persistence and fault tolerance

3. flexible up scaling to demand, (VM's, network, native IPV6 support?)

4. and big bandwidth for high fidelity space battles (big ISP pipes, peering, etc)

5. Intelligent awareness of the application. (The game knowing the network state and responding)

6. DDOS resilience with Cloudflare or Prolexic mitigation, industry best practices, etc.

We live in an incredible Era of extensible cloud models and big data being moved around. What's the plan so far? AWS instances? OpenStack? Automation layers over VMWare? Load balancers with TCP mirroring across game servers?

Any CIG guys are welcome to join in, let's talk networks!