Valkey 9.1 Adds Database-Level ACLs, JSON Logging, and a Redesigned I/O Threading Model for the Redis Fork
The Linux Foundation's Redis fork ships multi-tenant access control, native JSON logs, and throughput gains, cutting per-key memory by up to 10 percent.
Editor's Note ·
- Clarification:
- One cited source, diginomica.com (the diginomica.com/valkey-91-ships-hybrid-search-ai-maintainer-agents-engine article), is not on The Machine Herald's source allowlist. The outlet is an established enterprise-technology publication and every claim it supports in this article — the second Madelyn Olson quote and the durability/version-10.0 roadmap — was independently verified against the snapshot and corroborated by the primary Valkey blog and Linux Foundation press release. This note records the allowlist gap for transparency; it does not indicate a factual error.
Overview
Valkey 9.1, the latest release of the Linux Foundation’s open-source fork of the Redis in-memory data store, became available on May 19, 2026, bringing database-level access controls, native JSON-formatted server logs, and a redesigned I/O threading model. The release drew contributions from over eighty individual contributors, according to the project’s release blog. It continues the project’s effort to differentiate from Redis following the license change that prompted the fork.
What We Know
Security
Valkey 9.1 introduces a database-level access control system. According to the Valkey project blog, administrators and operators can now scope user permissions to specific databases. The Linux Foundation describes this as a database-level ACL system for fine-grained multi-tenant isolation within a single instance.
The release also moves the Lua scripting engine into its own module. By extracting Lua into a module, the project says it reduces the server’s security surface area. TLS handling improves as well, with automatic background reloading of TLS certificates and the ability to surface TLS certificate expiration dates via the INFO command, per the project blog.
Observability
For operators, Valkey can now emit server logs in JSON format with the log-format json configuration directive, making logs parseable by observability tooling without custom patterns. The release adds new cumulative metrics for main and I/O thread usage to help monitor server load.
Performance and efficiency
The release reworks the threading model. According to the Valkey project blog, a redesign of the I/O threading communication model improves throughput by up to 17 percent. The blog also reports up to 30 percent higher throughput for string GET commands, and notes that the XRANGE and XREVRANGE stream commands are up to 30 percent faster. Enabling the hardware clock by default reduces the overhead of time-related system calls, improving GET and SET performance by up to 3 percent.
On memory, the Valkey project blog cites up to a 20 percent memory reduction for strings under 128 bytes and up to a 10 percent reduction in sorted set memory usage. The Linux Foundation frames the broader efficiency gain as reducing per-key memory usage by up to 10 percent for common workloads with no tuning required, a characterization echoed by diginomica.
The release adds new commands including HGETDEL, MSETEX, and CLUSTERSCAN, per the project blog.
Ecosystem
Alongside the core engine, the project advanced several companion components. According to the Linux Foundation, Valkey Search reached version 1.2, Valkey Admin became generally available as an open-source visual cluster management tool, and Valkey GLIDE launched version 2.4 with new support for C# and PHP.
Madelyn Olson, Valkey project maintainer and Principal Engineer for AWS in-memory databases, framed the release’s emphasis on stability. “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,” she said, according to the Linux Foundation.
The project has also leaned on automation in its own development. Olson told diginomica that “Valkey has been using AI agents to automate tedious, time-intensive tasks now, while the wider industry is working on large-scale agentic models.”
What We Don’t Know
The GitHub release page lists Valkey 9.1.0 with an upgrade urgency of LOW, describing it as the first stable release of the 9.1 line, so the practical migration timeline for production operators is not yet clear. On durability, diginomica reports that full ACID compliance with fast failover is targeted for version 10.0, indicating that work remains outstanding beyond this release.
Analysis
Valkey 9.1 reflects the project’s continued positioning as a vendor-neutral alternative to Redis. As diginomica notes, Valkey is governed as a vendor-neutral project under the Linux Foundation, a contrast the project draws against Redis’s single-vendor stewardship. The 9.1 feature set, weighted toward multi-tenant security, native observability formats, and incremental performance and memory gains, is aimed squarely at the operational concerns of organizations running the data store at scale rather than at headline benchmark figures.