industryopinion
Gaming and data: Getting to grips with scale high performance gaming, from pixels to data
Patrick McFadin, Vice President Developer Relations at DataStax focuses on the future of data in gaming, and why real time data is still something that all gaming companies – from the smallest indie studio through to the biggest firms in the world – can struggle to get right.
T
he gaming sector has grown massively over the past two years. According to research by the University of Glasgow, 71 percent of people played more games during the
pandemic, while the frequency of gaming went up too - from around 10 percent playing daily to around 40 percent. Tis increase in play was a boon for the gaming sector, with revenues in the UK of £7billion in 2020, of which £4.55billion was sales of gaming soſtware. Tis growth in sales and play times has had a direct impact
on the needed infrastructure. In fact, gaming IT is critical to the companies involved as they support their players, when creating and saving player data, using social features, and analysing data to understand player behaviour.
Scaling up around data To support all these new players, gaming companies have had to scale up data management at the same pace. Tis data can be created at any time by huge numbers of players, and they all expect low latency and maximum uptime. Tis means fast match-making, little to no lag during play, and fair play experiences for everyone. Tis isn’t a simple task; it’s quite challenging to predict demand levels and capacity requirements. Scaling up usually means using cloud services, as it is more
flexible and cost effective than on-prem in most cases. Tis means a full scale migration of existing systems and data, or starting from scratch when planning a new game deployment. In either case, the primary concern is availability and performance. To get a head start on your project, you can look at what other
gaming companies have done in the past. Over the years, open source databases have been very popular with gaming backend systems as they can get you started and scale quickly without soſtware licensing cost. Apache Cassandra has been a leading choice, for its ability to scale up and support millions of players simultaneously. Tis database is in use at the likes of Activision, Blizzard, ESL Gaming, and Sony to support their massively online, real time services. Cassandra supports multiple deployment options, from
running your own cloud instance through to platforms like Database-as-a-Service (DBaaS) options. However, many developers want to have it all managed for them so they can
12 | March 2022
concentrate on game development rather than infrastructure. Serverless databases are rising in popularity for that reason.
Rather than needing to implement a pre-provisioned database or DBaaS platform, serverless database services work on demand and when needed. Developers only have to understand their APIs, and then use those to store and retrieve data. Tis approach provides capacity as it is needed, rather than based on estimates - by paying as you use, you can stop paying for over-provisioned systems, and, as a consequence, lower infrastructure costs. Te harder things like scalability are handled automatically
on the developer’s behalf. For example, say you have a game event launch that goes better than expected. Rather than being in trouble because you have not estimated capacity accurately, a serverless database will scale up in response to demand. You only pay for the exact amount of data that your players create over time, rather than over- or under-investing based on estimates beforehand. Your scaling costs track with your usage.
Getting more out of data Alongside storing data, there’s the opportunity to make more use of it by using streaming as part of your analytics approach. Tis can be used to spot trends in real-time data, or to flag activities for more investigation. Open source streaming services like Apache Pulsar can ensure
reliable stream data is where it is needed. Tis can maintain game infrastructure performance, user experience regardless of architecture changes over time. A single data point can also be used in multiple places, for
example fraud detection and security, where player behaviour is out of the ordinary. Taking this stream of data and looking for outlier behaviour can show up nefarious activity when its happening. Tis has to take place in real time, to stop players being taken advantage of. Te future for gaming is only expanding. For gaming
companies, data is an important asset that can be used to spot trends in their customer base and then respond. Cloud services and open source databases combined can help you scale, and serverless designs can make that process cost effective and automatic.
www.pcr-online.biz
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52
