Customer Success Story: Aternos

Who is Aternos?

Aternos is a leader in providing on-demand Minecraft game servers.

The company, founded in 2013, is providing services for over 400,000 Minecraft players daily and headquartered in Bonn, Germany.

Their Project

Aternos is running several hundred dedicated servers on-premises to provide backend services, as well as infrastructure for providing Minecraft instances. However, the capacity of dedicated hardware presents physical constraints resulting in long queues during peak hours. Therefore, they decided to expand their infrastructure onto cloud services.

The Solution

In 2019, more than one thousand virtual cloud instances were launched through Scaleway daily. Expanding their game servers into the cloud allowed them to significantly diminish the waiting time for a free slot, decreasing it from several hours to only a few minutes.

A Technical Point of View

The company runs a pool of several hundred on-premises servers, hosting the backend, storage, frontend, and game servers.

Administration and management of the game servers are mostly through PHP applications.

The Minecraft game servers run on-demand in Docker containers. Scaleway cloud instances are deployed dynamically depending on the workload on the platform with Ansible and AWX.

Aternos relies mostly on GP1-S instances as they offer the best performance/price ratio for their use case. They also deliver the performances needed with the least overhead possible. During peak hours, Aternos deploys up to 2000 cloud instances, each of them hosting several Minecraft game servers. The company is providing more than 40.000 Minecraft servers simultaneously at peak times, with two-third of them running in the cloud.

Scaling with Scaleway

“Thanks to Scaleway’s Virtual Cloud Instances and the exceptional customer assistance we experience, expanding our business onto the cloud was a smooth experience.” Explains Matthias Neid, CEO of Aternos.

Scaleway Cloud Services allow the company to deploy services on-demand when they are needed, without the hassle of deploying dedicated resources which are not in constant use. Resulting in lower operational expenses and business risks, better scalability, and increased performance and reliability.