Valkey 9.1 trims memory 10% and pulls search into the core
7 hours ago
- #database
- #open-source
- #performance
- Valkey 9.1 introduces a reworked internal data layout that reduces per-key memory usage by up to 10% for common workloads without requiring tuning or reconfiguration.
- The release adds security features like automated TLS certificate reloading and database-level ACLs for fine-grained multi-tenant isolation within a single instance.
- A new CLUSTERSCAN capability enables consistent key scanning across entire clusters, improving predictability for cluster operators under heavy load.
- Valkey Search 1.2 is now integrated into the core, offering full-text search, numeric filtering, tag-based lookup, and AI-ready vector search in a single system.
- Valkey Admin is generally available as an open-source visual cluster management tool for real-time deployment views, cluster topology, and key inspection.
- Valkey GLIDE client library version 2.4 adds support for client-side caching and expands language coverage to include C# and PHP, built on a shared Rust core.
- AI agents are being used to automate tasks like security scanning and backporting to combat maintainer burnout, while maintaining human review for core engine changes.
- Adoption is growing, with about 6 million weekly container pulls, a 17x increase from the previous year, and reported usage by large enterprises like Apple and Uber.
- The project positions itself as a long-term peer to Redis, aiming to compete and encourage more open-source contributions rather than replace Redis entirely.