Redis vs. Valkey: Navigating the Evolving Landscape of In-Memory Data Stores

In the fast-paced world of DevOps, choosing the right tools for your infrastructure can make or break your application’s performance and scalability. For years, Redis has been the undisputed king of in-memory data stores, a go-to for caching, session management, and real-time analytics. Its lightning-fast performance and versatile data structures have made it indispensable for countless developers and organizations. However, a significant shift in its licensing model has recently stirred the waters, leading to the emergence of a new player on the block: Valkey.

This isn’t just a tale of two databases; it’s a narrative about open-source philosophy, community stewardship, and the continuous pursuit of performance excellence. Let’s dive deep into what sets Redis and Valkey apart, helping you make an informed decision for your next project.

The Fork in the Road: Understanding Valkey’s Genesis

To truly grasp the distinction between Redis and Valkey, we must first understand the catalyst behind Valkey’s creation. For a long time, Redis operated under the permissive 3-clause BSD license, fostering a vibrant ecosystem and widespread adoption. This license allowed virtually anyone to use, modify, and distribute Redis, even within proprietary products, without significant restrictions .

However, in March 2024, Redis Ltd., the commercial entity behind Redis, announced a change in its licensing strategy. Starting with Redis 7.4, new versions would be released under a dual-licensing model, primarily using the Server Side Public License (SSPLv1) and the Redis Source Available License (RSALv2). More recently, with Redis 8.0, the GNU Affero General Public License (AGPLv3) was added as an additional open-source option. While these changes aimed to protect Redis Ltd.’s commercial interests and ensure sustainable development, they raised concerns within the broader open-source community, particularly among cloud providers and companies offering Redis-compatible services.

This concern stemmed from the restrictive nature of SSPLv1 and RSALv2, which placed conditions on commercial usage, especially for those providing Redis as a service. Many felt this deviated from the spirit of truly open-source software. In response, a consortium of prominent tech companies and former Redis core contributors — including Alibaba, Amazon, Ericsson, Google, Huawei, and Tencent — rallied together. Their objective was clear: to preserve a genuinely open-source, permissively licensed alternative. Thus, Valkey was born, forked from Redis 7.2.5 (the last version under the BSD license) and subsequently placed under the stewardship of the Linux Foundation.

This move ensures that Valkey will always remain under the BSD license, guaranteeing its unfettered use and distribution for all. It represents a commitment from a significant portion of the community to uphold the principles of open-source collaboration without commercial licensing constraints.

Under the Hood: Technical Deep Dive

While Valkey originates from Redis, its development trajectory is now distinct, leading to some notable technical differences and future aspirations. Let’s compare them across key technical dimensions:

Threading Model

One of the most significant architectural differences emerging between the two projects lies in their approach to threading:

• Redis: Historically, Redis has been largely single-threaded for command processing. This design choice simplifies concurrency control, avoiding complex locking mechanisms and ensuring predictable performance. However, for I/O-bound operations or particularly heavy workloads, the single-threaded nature could become a bottleneck. Recent versions, particularly Redis 8.0, have begun to introduce more multi-threading, specifically for I/O operations, to alleviate this.
• Valkey: From its inception as a fork, Valkey has aggressively pursued enhanced multi-threading capabilities. Valkey 7.2.5 and subsequent versions (like 8.0/8.1) feature an improved multi-threading architecture designed to boost throughput and reduce latency, especially for I/O-intensive tasks. This allows the main thread and I/O threads to operate concurrently, offloading tasks like reading and parsing network requests. Valkey boasts claims of achieving triple the speed and over one million requests per second (RPS) in certain configurations due to its multi-threaded advancements.

Performance and Scalability

Both Redis and Valkey are engineered for high performance, leveraging in-memory data storage to deliver low-latency operations. However, Valkey’s focus on multi-threading aims to push these boundaries further:

• Redis: Renowned for its speed, Redis excels in scenarios requiring rapid data access. Its single-threaded core, while efficient, means scaling often involves sharding data across multiple Redis instances.
• Valkey: With its advanced multi-threading, Valkey is designed to better utilize modern multi-core processors. This can translate to higher throughput and better scalability on a single instance, potentially reducing the need for extensive sharding in some use cases. It also offers more advanced observability with per-slot metrics, providing detailed insights into performance and usage patterns.

Ecosystem and Community

The ecosystem and community support are crucial factors for any open-source project:

• Redis: Benefits from a long-standing, mature ecosystem with a vast array of client libraries, tools, and integrations across virtually every programming language and platform. It has a large, established user base and commercial support from Redis Ltd.
• Valkey: As a newer fork, Valkey’s ecosystem is still nascent but growing rapidly. Its direct lineage from Redis 7.2.5 means it’s largely compatible with existing Redis client libraries and tools. The backing of the Linux Foundation and major industry players provides a strong foundation for future growth and ensures continued open-source development. Many of the original Redis developers and significant contributors have migrated to Valkey, bringing their expertise and dedication to the new project.

The Road Ahead: What This Means for DevOps

For DevOps professionals, the emergence of Valkey presents both a challenge and an opportunity. The choice between Redis and Valkey will increasingly depend on your specific priorities:

Licensing and Open Source Purity: If adhering strictly to permissive open-source licenses (like BSD) is paramount for your organization, or if you’re a cloud provider building services around an in-memory data store, Valkey offers a clear and unwavering commitment to this philosophy.

Performance and Multi-threading: For applications that demand the absolute highest throughput and can benefit from advanced multi-core utilization, Valkey’s aggressive pursuit of multi-threading might offer a performance edge.

Ecosystem Maturity and Commercial Support: Redis, with its established ecosystem and commercial backing from Redis Ltd., continues to be a robust and well-supported option, especially for those who prefer commercial agreements and a long track record.

Here’s a quick comparison table to summarize the key differences:

Conclusion

The landscape of in-memory data stores is evolving, and the introduction of Valkey marks a significant moment. While Redis continues its journey under a revised licensing model, Valkey stands as a testament to the enduring power of open-source collaboration. For DevOps teams, this means more choice and a renewed emphasis on understanding the nuances of licensing, community governance, and technical architecture.

Whether you stick with the battle-tested Redis or embrace the promising new path of Valkey, the core principles remain: leverage the right tool for the job, understand its implications, and continuously optimize for performance and reliability. The future of high-performance data infrastructure is certainly looking interesting!