Context

In a traditional datacentric backend system - we would need a database and a cache for performance. We would also require background process(es) which extract, transform and load data into the cache per business specs. This requires a high level of effort in maintenance and introduces complexity.

New Features of Redis

Redis is widely known as a caching database. However, it has evolved into a full blown primary database. Nowadays, it comes with an in-memory cache and a primary database in a single system. The ETL process(es) that sit between a database and cache are no longer required, reducing latency and complexity. Another feature is, Redis now provides multiple data models - Streams, RedisJSON, RedisGraph, RedisSearch - to name a few.

Benefits

The major benefit is the reduction in latency and complexity of a system. Technically, we no longer need ETL process(es) between database and cache. We may still need them for special cases, but the number of those processes will be significantly reduced.

With the multiple data model support, we can now set up Redis as a single source of data and build micro services with different business purposes on top of it.

End Notes

As a user of Redis, I find these very useful. And I am sure there are many more benefits and solutions that it may bring, such as in ML/AI-centric apps.

In next post, I will provide my tutorial on a simple experimental app that leverages Redis as a primary database and its different OM’s.