How I applied Redis to The Project

During my time at Gameloft working on Heroes of Order and Chaos, we encountered performance challenges with our player profile and matchmaking services. Our API response times were becoming slower as the user base grew, particularly during peak hours when thousands of concurrent players were accessing the system.

I identified this as an opportunity to implement Redis as our caching solution. While I hadn’t worked with Redis before, I took the initiative to learn it because its in-memory data structure store seemed perfect for our needs. I spent time studying Redis documentation, experimenting with its data structures, and testing different caching strategies.

I implemented Redis to cache frequently accessed data like player profiles, match history, and game configurations. I designed a caching strategy that included:

• Setting appropriate TTL (Time To Live) values for different types of data
• Implementing cache invalidation patterns to maintain data consistency
• Using Redis’s sorted sets for real-time leaderboards
• Leveraging Redis pub/sub for real-time game events

As a result, we saw average API response times decrease by 60%, and our database load was reduced by approximately 40%. The system became much more resilient during peak hours, and we were able to handle sudden traffic spikes more effectively. This experience taught me that staying open to learning new technologies and carefully evaluating their potential benefits can lead to significant improvements in system performance.