The Linux Foundation has announced the general availability of Valkey 9.0, the open-source in-memory storage solution developed as a successor to Redis. The latest major version introduces atomic slot migrations, hash field expiration, and full support for numbered databases in cluster mode, enabling scaling to 2,000 nodes and achieving over 1 billion requests per second.
Released one year after Valkey 8.0, Valkey 9.0 adds atomic slot migration, improving how clusters rebalance data. Unlike stepwise migrations that could change ownership mid-transfer, the new atomic movement is designed to ensure consistent key routing and predictable handoffs, reducing transient errors and simplifying live resharding. Khawaja Shams, co-founder and CEO of Momento, and Allen Helton, ecosystem engineer at Momento and AWS Hero, write:
For teams running Valkey in clustered environments, this fundamentally shifts how you plan capacity and manage operational risk. Scale-outs become predictable instead of painful.
Kyle Davis, senior developer advocate at AWS and open source leader for Valkey, explains:
In Valkey, keys are bundled into one of 16,384 ‘slots’ and each node takes one or more slots. In Valkey 9.0 instead of being key-by-key, Valkey migrates entire slots at a time, atomically moving the slot from one node to another using the AOF format.
Another prominent feature is hash field expiration. Previously, Valkey hashes could only expire as a whole, forcing users to split data across multiple keys when field-level expiration was needed. Valkey 9.0 allows individual hash fields to expire independently. In a separate article, Ran Shidlansik, senior software engineer at AWS, explains how hash field expiration works and why active, rather than lazy, expiration is used to reclaim expired hash fields. Shidlansik concludes:
The benchmarks demonstrate that field-level expirations can be added to Valkey without compromising memory efficiency, or latency. The memory overhead remains modest and predictable, command throughput is unaffected, and the shared active expiration job efficiently reclaims memory even under heavy ingestion workloads.
Numbered databases, long used to separate data and prevent key collisions, were limited in Redis and previous Valkey releases to a single database in cluster mode. Removing this restriction, Valkey 9.0 introduces full cluster support for numbered databases, enabling scalable, multi-database deployments. Davis explains how numbered databases are a form of namespacing and their primary use case:
The most straight forward use case of numbered databases is when you need to separate your data logically and you can tolerate the effects of resource sharing. This might be something like keeping customer data separated from one another or combining applications on to a single cluster when resources are unlikely to be an issue.
Source: Valkey blog
The community behind the project recently discussed the architectural improvements in Valkey 9.0, which demonstrated how the resilience of large clusters enabled scaling to 2,000 nodes and achieving over 1 billion requests per second. Shams and Helton add:
The performance story in 9.0 is about using modern CPU capabilities intelligently. Valkey 9.0 takes advantage of modern hardware to deliver performance that outpaces previous versions by wide margins (…) As a result, you get lower tail latency, higher per-node throughput, and measurable cost efficiency at scale.
Supported by AWS, Oracle, and Google Cloud, among others, Valkey is released under the BSD 3-clause license. Valkey 9.0 is now available for download.
