Only as good as the data center behind it: Scalability and load management in high-quality online gaming

In 2006, Blizzard Entertainment, the company that owns World of Warcraft (WoW), launched a worldwide video gaming event so ambitious that to date—thirteen years later—the experience still makes every hard-core gamer’s top-ten list.

The “Opening of the Gates of Ahn’Qiraj” was designed as a one-time-only WoW war event that required cooperation on a scale never seen before. Hundreds of players from both the Alliance and Horde factions came from every corner of the game’s world map, gathering supplies, running down quests, and fighting monsters in what would become a months-long war to open the gates and unleash the monsters and bounty behind them.

The problem was, to open the gates, all players would inevitably converge on a single area all at the same time. This, in hindsight, was not a good plan.

At the event’s climax, servers in WoW’s seventeen data centers converged in one place simultaneously, maxing out CPUs and databases and proceeding to repeatedly crash the entire network of world servers over and over and over again until players couldn’t stay awake any longer. Only when enough people dropped off to lower the threshold could a player finally open the gate—a frustrating end to what had been an epic journey.

“Turns out that you should totally be talking to your server engineers when you have a plan like this,” said Alex Afrasiabi, WoW’s creative director, in the documentary World of Warcraft: Looking for Group. “We just didn’t realize how many people would actually show up to this thing.”

WoW lived to see another day. Blizzard reported Q2 2018 earnings for WoW of more than $1.1 billion. But it’s safe to say that when it launches World of Warcraft: Classic this summer (including a 2.0 version of the “Opening of the Gates of Ahn’Qiraj”), the tolerance for stuttering or frozen screens will be zero.

Today, massively multiplayer online role-playing games, or MMORPGs, like WoW and Fortnite, the zombie survival game, require ten to forty megabytes per hour to play, but the numbers can rise quickly. For example, Fortnite depends on twelve data centers that have to manage peak loads that are ten times larger than the smallest ones.

Gaming has become a multi-billion-dollar industry where the choices players make dictate demand. Modern data centers need the capacity to reliably manage regular load fluctuations. Games need to be up and running all the time, anytime, with a seamless, high-quality user experience.

ROOT Data Center’s high-density power solutions provide the performance, scalability, and reliability the gaming industry requires. It has the capacity to deploy enterprise customers in less than twenty-four hours, allowing clients to scale up quickly as their growing business demands. Additionally, ROOT delivers 100 percent uptime, which is important to game production teams that require an “always on” infrastructure for development and testing of their products to meet competitive speed-to-market challenges.

Did you know that Montreal is the fifth-largest gaming center in the world, meeting needs for not just incentives, talent, and space but also for IT infrastructure? To learn more, download ROOT Data Center’s paper, “Why Montreal Data Centers Are the Best Choice for Video Games.”