Valkey 9.0 Released: Faster Clusters, New TTL Controls, and Big Networking Gains

Image source: https://www.linuxfoundation.o…

TL;DR

Valkey 9.0 debuts with atomic slot migrations, hash field expiration, and improved cluster mode support, enhancing data management and scalability.

Key Points

Highlight key points with color coding based on sentiment (positive, neutral, negative).

Valkey 9.0 introduces atomic slot migrations, enhancing data migration efficiency and reducing performance issues.

Hash field expiration allows developers to set expiration times for individual fields, offering more granular data lifecycle control.

The update expands support for numbered databases in cluster mode, facilitating better data separation and scalability.

Performance enhancements include pipeline memory prefetching and SIMD optimizations, contributing to higher throughput.

Scalability improvements enable the system to handle over 1 billion requests per second with large clusters.

Valkey 9.0 is here, and it's bringing some interesting changes to the table, especially for those of us knee-deep in data management. One of the most exciting updates is the introduction of atomic slot migrations. Remember the days of moving keys one by one? Yeah, that was a bit of a hassle. Now, Valkey lets you move entire slots at once. This nifty feature not only sidesteps the old headaches of performance dips and data loss risks but also keeps your data accessible without those pesky client retries during migrations.

There's more! Let's explore hash field expiration. In the past, you could only set expiration at the key level, which was a bit limiting. Developers had to jump through hoops to manage data efficiently. With Valkey 9.0, though, you can set expiration times for individual fields within hash data types. New commands like HEXPIRE and HSETEX make it possible for specific fields to expire on their own. This change is a significant improvement for optimizing memory usage and reducing operational complexity.

And if you're working in cluster mode, there's good news. Valkey 9.0 now supports multiple numbered databases. Previously, you were stuck with just one, which could lead to key name clashes. This update offers more flexibility and helps avoid those annoying conflicts.

There's a lot more under the hood, too. For large clusters, Valkey 9.0 can handle over a billion requests per second, thanks to pipeline memory prefetching that boosts throughput by up to 40%. Plus, 25 previously deprecated commands are making a comeback. Zero-copy responses for large requests, Multipath TCP to cut down on latency, and optimizations for SIMD in BITCOUNT and HyperLogLog operations are all part of the package. With new features like conditional delete and CLIENT LIST filtering, Valkey 9.0 is clearly focused on performance and usability.

Key Numbers

Present key numerics and statistics in a minimalist format.

2,000 nodes

The maximum number of nodes Valkey 9.0 can scale to.

1 billion requests per second

The maximum number of requests per second Valkey 9.0 can handle.

40 %

The increase in throughput due to memory prefetching when pipelining.

20 %

The increase in throughput due to zero copy responses.

200 %

The increase in throughput due to SIMD for BITCOUNT and HyperLogLog optimizations.

25 %

The reduction in latency due to Multipath TCP support.

25 %

The maximum percentage of available CPU time used by the active expiration job.

Stakeholder Relationships

An interactive diagram mapping entities directly or indirectly involved in this news. Drag nodes to rearrange them and see relationship details.

Organizations

Key entities and stakeholders, categorized for clarity: people, organizations, tools, events, regulatory bodies, and industries.

Valkey Open Source Project

Released Valkey 9.0 with major performance, scalability, and data management improvements.

Tools

Key entities and stakeholders, categorized for clarity: people, organizations, tools, events, regulatory bodies, and industries.

Atomic Slot Migrations Cluster Migration Feature

Moves entire slots at once to make data migration faster and more reliable.

Hash Field Expiration Data Lifecycle Feature

Allows expiration at the individual hash-field level for more fine-grained TTL control.

Numbered Databases in Cluster Mode Cluster Feature

Restores support for multiple numbered databases in clustered configurations.

Pipeline Memory Prefetching Performance Optimization

Improves pipelined command throughput by reducing memory stalls.

Multipath TCP Support Networking Improvement

Adds support for MPTCP to improve reliability and performance across multiple network paths.

Events

Key entities and stakeholders, categorized for clarity: people, organizations, tools, events, regulatory bodies, and industries.

Valkey 9.0 Release Product Release

Announcement of new features, performance enhancements, and cluster improvements.

Timeline of Events

Timeline of key events and milestones.

Sep 30, 2025 Introduction of hash field expiration in Valkey

Valkey introduced the ability for individual field expiry within hash data types.

Oct 21, 2025 Announcement of Valkey 9.0

Valkey 9.0 was announced, featuring innovations like atomic slot migrations, hash field expiration, and expanded support for numbered databases in cluster mode.