> ## Content Index
> Fetch the complete content index at: https://www.stirup.si/llms.txt
> Use this file to discover other available public pages before exploring further.

# Olson and Söderqvist Discuss Valkey’s Evolution and Future Past Caching Use Cases at OSS EU
- URL: https://www.stirup.si/olson-and-soderqvist-discuss-valkeys-evolution-and-future-past-caching-3add582a/
- Published: 2026-10-12T05:17:07.000Z
- Updated: 2026-10-12T05:17:07.000Z
- Description: Valkey, a fork of Redis launched over two years ago, has evolved significantly with improvements in throughput, cluster functionality, and plans for features like synchronous replication and tiered storage. The project emphasizes security and release efficiency through automation.
- Author: Satish Das
- Tags: Development, Stirup Research

## Valkey’s Technical Evolution and Strategic Direction

Valkey, a fork of Redis initiated over two years ago, has grown into a project with a strong foundation and a clear roadmap. Supported by twelve maintainers and nine technical steering committee members from companies such as Ericsson, AWS, Apple, and Percona, Valkey has made significant strides in performance and scalability. During the Open Source Summit Europe, two of its initial developers, Madelyn Olson and Viktor Söderqvist, provided insights into the project’s evolution and future directions.

The 8.x releases of Valkey focused on foundational infrastructure improvements. One of the most notable changes was the introduction of async I/O threading, which increased throughput from approximately 200,000-250,000 requests per second to around one million per process. This enhancement allows operators to dynamically adjust thread configurations at runtime, enabling better handling of burst workloads. Additionally, memory efficiency was improved through key embedding and the use of per-slot dictionaries. Dual-channel replication was also introduced, which reduces memory pressure during replica synchronization by sending dataset copies and replication changes in parallel.

The 9.x releases addressed significant limitations in cluster mode. Previously, cluster mode had two major constraints: cross-slot operations required data to reside on the same shard, and it did not support numbered databases, which are available in single-instance deployments. The introduction of numbered databases removed one barrier to migration to cluster mode, while atomic slot migration ensured that data migrations either complete fully or not at all, preventing partial migrations in case of failures. Hash-field expiration, a highly requested feature, was also added in version 9.

The GLIDE client library was developed to simplify client behavior in complex environments. Originally, Redis clients had to handle cluster topology, TLS, redirects, and connection pooling, which led to inconsistencies across different language implementations. GLIDE centralizes these functions in a shared Rust core, allowing for more consistent and efficient client behavior across languages. While WebAssembly was considered, the team opted against it at this stage, and GLIDE is not intended to replace native clients but to complement them.

Looking ahead, Valkey aims to expand beyond its current caching use cases. Synchronous replication is being explored to provide durability guarantees for non-ephemeral data. Tiered storage is another area of interest, with the potential to keep hot data in memory while moving colder data to SSD or other storage solutions. The team is also considering a more general interface that could support backends beyond local NVMe, such as S3 or Postgres. Cross-cluster active-active replication with eventual consistency is a longer-term goal, though not an immediate implementation commitment.

Beyond feature development, Valkey is also investing in release and security improvements. Automated backporting is being used to streamline the process of applying fixes across different releases, while AI-assisted pre-review and proactive security testing are being implemented to identify potential issues before a release.

**Source:** [InfoQ](https://www.infoq.com/news/2026/10/valkey-evolution/?utm%5Fcampaign=infoq%5Fcontent&utm%5Fsource=infoq&utm%5Fmedium=feed&utm%5Fterm=global). Analysis based on publicly available information; no hands-on testing is claimed.