What is Redis?

Data latency is the amount of time between data events, such as the time between generation and availability for a query or the time between the query and delivery to the application.

Latency is typically measured in milliseconds. However, real-time applications and analysis require as low of latency as possible, and benefit from the microsecond latency times offered by in-memory databases such as Redis. 

Valkey is an open source database that is a fork of Redis. It was formed after Redis Inc., which used to distribute Redis under an open source license, adopted a "source available" license for Redis in 2024. To maintain an open source alternative, several core Redis community members launched Valkey, starting with the last open source version of Redis. Generally speaking, the technical information about Redis on this page is equally applicable to Valkey.

The structure of a database has a significant impact on data latency. Redis is a NoSQL database. NoSQL databases are non-relational databases that store data in a variety of data models, including document, key-value, graph, and columnar formats. They’re flexible, can handle large unstructured datasets, and are designed for fast search queries. Furthermore, Redis is an in-memory database, which means it can offer low-latency reads and writes. For applications that require low latency, a NoSQL, in-memory database like Redis or Valkey is the best choice.

Redis databases are:

• Fast. In-memory databases don’t require disk access, reducing the latency from milliseconds (ms) to microseconds (μs).
• Flexible. NoSQL databases store data without a fixed schema—Redis databases inherit this flexibility advantage.
• Scalable. The structure of a Redis database involves splitting and distributing data across multiple databases or machines, a process known as "sharding"—enabling horizontal scalability. 
• Simple. Redis enables developers to write potentially complex code in fewer lines, so it’s accessible for new users to learn and use.
• Highly available. Automatic failover is available, so applications remain operational even during failures.
• Open source. Older versions of Redis and all versions of Valkey are open source, meaning that you can use those versions for your applications with no licensing cost and full flexibility in deployment. Newer versions of Redis are offered on a "source available" basis, and may require a license.

Managed databases are cloud-based services where the cloud service provider takes care of maintenance, security, patching, and provisioning. This contrasts with a self-managed database, where your developers or IT team are responsible for these tasks.

You can choose to self-manage your Redis or Valkey implementation; however, there are limitations and tradeoffs that can cause complications if the database is not managed properly. For example, scaling is more complicated with a self-managed Redis database, as you will need to add servers manually.

A cloud-based managed service provides lower cost and peace of mind, removing the burden of managing your deployment. A fully managed Redis or Valkey solution provides worry-free scaling, availability, and reliability, freeing developers’ time for more strategic tasks like data modeling, performance optimization, and deriving value from the data.