Percona Announces Valkey-proxy to Ease Redis to Valkey Migration

RALEIGH, N.C., Oct. 07, 2026 /CourierPR/ -- Percona, a provider of open source database software, support, and services, has announced the launch of Valkey-proxy, a new open source project designed to ease the migration of enterprises from Redis to Valkey. This project, developed by Percona engineers, will provide a proxy layer that enables applications built for a single, standalone Valkey instance to run across a horizontally scaled-out cluster.

Valkey-proxy will allow teams on commercial Redis to have a clear path to migrate and simplify communication between instances for teams already running Valkey. The project will sit between the application and Valkey instances, manage communication between them, and present a cluster to the application as if it were a single standalone deployment. It also pools backend connections, ensuring connection counts remain bounded as applications scale.

Valkey is a high-performance, in-memory key-value data store used for caching and for database workloads where speed is essential. It runs in two modes: standalone mode, which suits smaller deployments with redundancy provided by failover replicas, and cluster mode, which is built for larger deployments with scale and redundancy designed in from the start.

Teams that start on standalone deployments often rely on capabilities that behave differently in cluster mode, such as multi-key commands, keyspace scanning, and pub/sub. Cluster mode also requires every client to track which node holds which data and handle redirects as the cluster changes. Moving to a cluster has typically meant rewriting parts of the application. Valkey-proxy removes that step by providing a proxy layer that manages communication between instances and presents a cluster to the application as if it were a single standalone deployment.

Development is underway, and the code will be published in mid-October under the BSD-3-Clause license. A public feedback and testing period will follow, with general availability targeted for early 2027. The goal is to give companies a path to move complex Redis deployments to a fully open source stack with minimal changes to their applications.