Presented by i3D.net's Game Hosting Platform
The rising popularity of multiplayer games continues to play a major role in the rapid growth of the video game industry. We are seeing developers from AAA studios to indie developers move towards a more multiplayer-centric portfolio of games. This expands the scope of work for developers and publishers, meaning that the core business of developers and publishers moves from "just" making and publishing games to now also including the support of a live service. If you’ve decided to not spend top dollar on flexible cloud solutions, nor want to overcommit on dedicated resources, it is likely you’ve arrived at the conclusion that you need game server tools to scale your compute resources up and down based on real-time concurrent users (CCU).
Building your own game server management tools
We probably don’t need to explain that building scalable game server tools is a complex and challenging task that requires a significant investment of time and resources. If you’re part of a large and established studio or publisher, this task is likely to be made easier by access to existing scaling tech and tools, as well as colleagues with years of experience in building and managing the tech needed to successfully create your own. However, for independent game studios, the lack of access to this technology and expertise in-house will make this process much more difficult.
Aside from whether you have access to the tech and expertise in-house, you’ll need to consider at least three important points that may be easily overlooked:
It needs to be secure and stable with minimal downtime. This becomes a wholly different concept when games are the service in question: many players play 24/7, so the tech you build will have to be secure and stable around the clock as well. If the studio has a very small team that works on this, they will also be on standby whenever something happens (and something always happens).
The tech you build would also have to handle scaling up and down for spikes in player traffic very quickly. That is the whole point of building scalable multiplayer game server tools. But this can be very difficult to do depending on the strategy in scaling up. If you are scaling up into one cloud provider, what if the location and instance type you want to scale up into is sold out? What logic are you implementing into your tech to make sure it can always scale up for player spikes, regardless of the region or what is happening with your vendors, as you’d still be dependent on them for compute resources?
Privacy policies, GDPR, and the constant adding of new features to comply with changing regulations, new game engine versions, game builds, etc are other factors to consider. As mentioned above, running a live service or any multiplayer game is 24/7. Much like owning a dog, you won’t be leaving the house anymore.
Choosing the right provider
It’s good to see many people realize how difficult it is to build scalable multiplayer game tech, as the people that have experience with it either started their own companies creating scaling tools, or work with infrastructure companies that have integrated automated scaling products into their product portfolio. There are various ways to approach the challenge of multiplayer game development, depending on your specific needs and requirements. One way to categorize potential solutions is by the type of provider offering them.
Infrastructure providers offer compute and connectivity resources to support multiplayer games. Some infrastructure providers that focus on the video gaming industry also have software build tools that help manage online multiplayer services, such as a matchmaker.
Orchestrator software providers offer tools that take care of some of the management of the infrastructure aspect, making it more hands-off for the developer. However, these providers do not own the actual infrastructure the game runs on, which can make it challenging to resolve issues and guarantee low latency.
Backend as a service (BaaS) providers go beyond infrastructure solutions and offer a broader range of solutions throughout the game development lifecycle. However, they also come with similar challenges as the orchestrator software providers and may impact user experience.
Whichever type of provider works best for you depends again on your expertise and the tools available. But as always, there are two important variables to think about prior to deciding: a) How could my decision impact player experience? and b) How could my decision impact cost control during the lifecycle of my video game? The last thing you want to do is have to migrate everything to another solution to make sure your player base can play the game or your costs are not astronomical when your game goes past the growth phase.
If, after calculating the cost for multiplayer infrastructure with your expected CCU, you’ve come to the conclusion that going 100 percent with on-demand cloud may be hurting your bottom line, or that committing 100 percent of your resources in case of a CCU spike may not be the most efficient strategy, you’ll probably need some sort of game server tech. A solution that can automatically scale resources up and down based on CCU and preferably one that predicts CCU in a way that you can scale up right before players join, so players do not get annoyed by waiting or bad servers, would be ideal.
In the hands of experts
Due to its focus on the video gaming industry, and seeing the difficulties described in this article, i3D.net developed software that would make the management of multiplayer titles easier. One of those products is its new Game Server Orchestrator, which runs your game server resources based on CCU. This helps you manage your game servers as efficiently as possible, minimizing overhead while prioritizing player experience by reducing potential wait times as a result of server overloads.
The orchestrator spins up game sessions and instances automatically based on the demand, and does so both on i3D.net compute resources, which it controls fully in order to ensure the quality of the network and compute, but it can also burst game sessions into compute resources not owned by i3D.net such as major cloud providers like GCP, Azure and Amazon Web Services (AWS) to ensure that when your game blows up, it can launch game instances in multiple regions instantly, across multiple providers to keep your community online. The orchestrator is complemented by the integration of i3D.net’s Content Delivery Network (CDN) for game build distribution, a system to patch game builds without impact for your player base as well as an Anti-DDoS solution that adds no latency for your user to keep your game servers safe and your community online.
As an infrastructure provider, i3D.net offers compute and connectivity resources to support multiplayer games. The company started out as a game hosting company and therefore its entire infrastructure stack is optimized for multiplayer games to ensure low latency. Its network is designed to reach global internet hubs and the hard-to-reach edges of the network alike, ranked the 14th most connected network in the world, ensuring users across the globe can be connected seamlessly.
If you are interested in finding out more about the Game Server Orchestrator and how it could help manage compute resources for your next multiplayer project, reach out to our team members at D.I.C.E, Las Vegas on February 21-23, 2023, or the Game Developers Conference (GDC), San Francisco on March 20-24, 2023.