Live Tech Games

The Challenge

Live Tech Games needed a tailored gaming platform that could handle millions of players, present unique game requirements and workload challenges. The company saw the need to find an enterprise game platform that could support millions of players with little maintenance cost but would also require it to be cost-effective for their business model to work.

One of the biggest challenges is scaling and constructing a system that can support 2 million realtime users for 10 minutes a day while keeping the server cold the rest of the day.

Several obstacles to overcome along the road, including dealing with unstable connections or having those same connections dropped after being established again as they played matches against each other. 

The most difficult aspect of organizing this tournament was determining how to handle when matchups were initially given and then gradually filtered out, with a final game and winner. 

The Solution

We've been to cut down on the costs associated with scaling our architecture and being efficient in scaling services experiencing a high volume of traffic. Advancing from microservice architecture helps us scale just those services that need it most so we can avoid unnecessary overhead and cost without compromising performance.

Using .NET Core in conjunction with the SignalR software library, we overcame the primary problem of transitioning between distinct game states/events. We upgraded to the Azure SignalR service so that our system could expand better and support a million concurrent users. This gave us a massive advantage on many other games that have struggled to keep up with abrupt increases in player traffic. Microsoft Azure Kubernetes has been used to orchestrate services across the cluster.

The automation tests covered a wide range of scenarios involving player interactions with one another and the system as a whole.