Valkey 9.1 cuts memory use & boosts security tools

Valkey has released version 9.1 of its open source key-value database, adding updates to security, memory efficiency and operational tooling.

The release was shaped by more than 80 contributors and arrives alongside updated ecosystem tools including Valkey Admin, Valkey Search 1.2 and Valkey GLIDE 2.4.

Version 9.1 changes how the database stores and manages data internally, cutting per-key memory use by up to 10% for common workloads without requiring configuration changes. For operators running large deployments, that could reduce infrastructure spending while maintaining the same application footprint.

Security is another focus. The release introduces automated TLS certificate reloading and a database-level access control list system designed to give administrators more precise isolation for multiple tenants within a single instance.

It also adds CLUSTERSCAN, a feature that lets operators scan keys consistently across an entire cluster. That addresses a common management task for teams running distributed deployments and expands the tooling available for inspecting production environments.

"Our goal as a project is to ensure that Valkey delivers new functionality while behaving very predictably - specifically we want to sustain our best-in-class performance and efficiency," said Madelyn Olson, Valkey Project Maintainer and Principal Engineer for AWS in-memory databases.

"We're able to consistently achieve this by continually investing in the modularity of the engine, which enables new functionality like full text-search," Olson added.

Search module

Alongside the core database release, the Valkey community has updated Valkey Search to version 1.2. The module combines full-text search, numeric filtering, tag-based lookup and vector search with the same system used for data storage, reducing the need for a separate search platform.

That approach reflects a broader push among infrastructure teams to simplify data stacks that often rely on multiple specialised systems for caching, retrieval and indexing. By bringing those functions closer together, organisations can reduce the number of moving parts involved in supporting real-time applications such as eCommerce sites, dashboards and leaderboards.

The search module can scale to terabytes of data while handling large request volumes. Its vector search support also places the project within a growing class of databases and add-on modules being adapted for AI-related workloads, especially systems that need semantic retrieval alongside conventional search.

Admin tools

Valkey Admin has reached general availability as an open source visual cluster management tool. It gives operations teams a live view of cluster topology, per-node performance and stored keys, while also exposing hot key detection and logs for slow or large requests.

Those features are intended to reduce reliance on custom scripts and command-line access for day-to-day diagnosis. The tool also includes sampling and rate-limiting controls so teams can inspect active systems with less risk of adding overhead.

For engineers managing distributed databases, visibility into hot spots and request behaviour is a recurring operational issue. A built-in management interface could make Valkey easier to use for teams that do not want to build their own dashboards or inspection tools around the database.

Client update

Valkey GLIDE, the project's client library, has been updated to version 2.4. The release adds support for client-side caching and transparent caching, while extending language support to C# and PHP.

The library is built on a shared Rust core and is designed to provide consistent behaviour across multiple programming languages. It also handles cluster topology, connection management and request routing automatically, reducing the amount of code application teams need to write to work with distributed deployments.

Client-side caching allows frequently requested data to be served from local memory instead of being fetched repeatedly across the network. That can be particularly useful for read-heavy applications in which the same keys are requested at very high rates, because it reduces repeated round trips to the database and lowers server load.

Valkey sits under the Linux Foundation and is backed by contributors and industry participants including Aiven, Alibaba Cloud, AWS, Bytedance, Ericsson, Google Cloud, k0rdent, NetApp Instaclustr, Oracle, Percona and UpCloud. The project supports workloads including caching, messaging and primary in-memory storage under a BSD licence.

The 9.1 release shows how the community is working to make the database more modular while expanding the surrounding tools used to operate it.