Getting your database right is critical for unforeseen digital challenges

16 September 2020 | Ronald Hoogland, Head of IT, Gamepoint

Cover

With the past coronavirus lockdown affecting all aspects of our life from how we worked to how we played, it was no secret that many looked to gaming as an escape.

According to Verizon, in the US alone there was an 82% increase in video game usage during lockdown compared to pre-COVID levels, as gamers sought to entertain themselves while confined to their own homes. In the early weeks of the lockdown, this led to some games and gaming platforms smashing their usage records. Counter Strike: Global Offensive saw over one million concurrent players, a record for the eight-year-old game, and video game distribution service Steam reached a new record of 20.3 million concurrent users online.

For the most part, the gaming industry stepped up and offered its own special contribution to  easing the effects of the pandemic on those under lockdown. However, many platforms both big and small were pushed to breaking point under the weight of increased user demand. Call of Duty, FIFA and Grand Theft Auto Online all suffered outages, with the latter two - both of which are among the world’s most popular games - even forced to take their services offline for hours, much to the frustration of players.

Whilst these times have undoubtedly been exceptional, gaming platforms should heed the lessons from the past lockdown and ensure they have the right technology in place to deliver reliable, high performance at scale with continuous availability across their platforms, and in doing so maintain their customers’ satisfaction.

The Gamepoint story: Continuity holds the key

At Gamepoint, a multiplayer social gaming network founded in the Netherlands, we too saw a significant surge in usage as the threat from coronavirus grew. During the early lockdowns our systems saw a 20% increase in average minutes played by all of our players, with a 10% increase in the number of concurrent users who were active on our platform.

The majority of our players turned to our platform first and foremost for gaming and entertainment, but to many, it was able to provide much more to help them connect to the world. Thanks to our in-chat function, all our games have developed their own global communities in which our players socialise and communicate on a regular basis.

It  quickly became apparent to us that a large volume of our players were using the platform and chat room as an important forum for social interaction outside their home. It was therefore paramount that we needed to keep our platform available through a period of high demand to support these individuals’ wellbeing and their ability to socialise during such a turbulent period.

Our ability to maintain fast, scalable access through our data layer proved a central factor to the continuity of our platform throughout this period of increased demand. By forming a strong working partnership with Redis Labs, we were able to maximise our uptime and performance to ensure the service remains operational throughout government lockdowns and shelter-in-place orders without any hitches.

Ensuring high performance and availability

It’s no secret that the strength of the data layer as part of the overall architecture underpins any application. More often than not it provides the foundations on which great applications and services are built.

Historically we’d experimented with different types of databases, such as MariaDB, to try and find the right bedrock for our offering. Whilst MariaDB worked well as a MySQL database, it proved incapable of handling certain datasets, which had a knock-on effect on performance. To counter this issue we decided to move to Redis five years ago. Using an open source database gave us the flexibility to choose the right solution for our needs, and in time we eventually upgraded to its commercial edition, Redis Enterprise, which provides us with operational hardening to help with scaling our systems. Throughout this period we were also able to get an understanding of what data we could handle and how it could be moved, which helped to improve our overall performance and throughput.

Redis Enterprise also turned out to be the perfect choice in light of the unexpected peak in usage we saw due to the coronavirus pandemic.

As our user numbers have continued to grow, Redis Enterprise has allowed us to adapt to increased usage and help deal with the additional session and caching requests perfectly, flawlessly managing the increase in volume of database requests with great speed, whether for leaderboarding, in-game chats or tracking player sessions.

We’re proud to say that we saw 100% uptime during the pandemic, a large part of which was down to the rock solid reliability of Redis Enterprise. Getting the data layer performance right helped us serve players with the instant experiences they demanded and nurture their loyalty, which included increased purchases of our virtual currency within our platform.

Ensuring a seamless user experience during the period has proven pivotal to our plans to continue growing and expand the platform, and it’s great to know that we’ve been able to play our part in keeping our users socially engaged. Partnering closely with strategic vendors has provided us with the foundation to keep our systems running, and keep our players connected. We certainly aren’t hoping for anything like the pandemic to come along again soon, but it has provided the live-fire drill we needed to test our platform and give us confidence that our systems are ready for any challenges that lie ahead.