readysettech/readyset
Readyset
Readyset is a transparent caching proxy for PostgreSQL and MySQL that sits between an application and its database, intercepting SQL queries and serving cached results from memory. It automatically caches query results on first execution and keeps those caches consistent by consuming the database’s replication stream in real time, enabling faster repeated reads without application code changes. The proxy also supports caching advanced SQL functions such as window functions, bucket functions, and locale-aware collation sorting, and exposes an interface that allows AI agents to inspect proxied queries and dynamically create or drop caches.
What distinguishes Readyset is its real-time cache synchronization via change-data-capture, combined with high-availability support through global transaction identifiers (GTIDs) to maintain consistency during database failovers. Developers can embed SQL comment hints to control caching behavior on a per-query basis, and a CLI tool allows diagnosing SQL workload performance before production deployment. The proxy also features agent-oriented cache management, where AI agents can programmatically inspect workloads and adjust caching strategies.
Beyond its core identity, Readyset offers a range of operational capabilities: it can be deployed as a transparent proxy without modifying application code, installed as a binary or packaged as systemd services for x86-64 and ARM64, and integrated with existing connection poolers and cloud platforms. Monitoring features list all cached SQL queries and their status, while cache management is also available through custom SQL commands to create, show, or remove cached queries.
Installation options include a downloadable binary, deb and rpm packages for systemd service deployment, and setup that does not require changes to application code or ORM layers.
Features
- Database Proxy Caches - Acts as a transparent caching proxy that intercepts SQL queries for PostgreSQL and MySQL databases.
- Cache Management Interfaces - Exposes server endpoints for AI agents to inspect workloads and manage cache states.
- Change Data Capture - Consumes the database replication stream to keep cached results consistent in real time.
- Query Result Caching - Caches SQL query results in memory and automatically invalidates them on data changes.
- In-Cache Query Execution - Automatically caches SQL query results on first execution and keeps them consistent by consuming the database replication stream in real time.
- Database Traffic Interception - Intercepts database protocol traffic transparently without requiring application code changes.
- AI-Driven Cache Controllers - Ships an interface for AI agents to inspect queries and dynamically create or drop caches.
- High Availability Cache Layers - Ensures cache consistency and high availability during database failover using global transaction identifiers.
- Analytical Result Caches - Caches advanced analytical functions such as window functions, bucket functions, and locale-aware collation.
- Selective Query Caching Markers - Allows marking specific SQL queries for proxy-caching via embedded SQL comment hints, enabling selective caching.
- SQL-Based Cache Management Commands - Enables cache management via custom SQL commands to create, show, or remove cached queries directly.
- Advanced SQL Function Caching Support - Supports caching window functions, bucket functions, and locale-aware collation sorting for complex analytical queries.
- SQL Performance Analysis - Provides a CLI to analyze and diagnose SQL workload performance before production deployment.
- Cache Directives - Allows developers to control caching behavior per query using SQL comment hints.
- GTID Consistency Mechanisms - Uses global transaction identifiers to maintain cache consistency during database failovers.
- SQL Workload Analysis Tools - Provides a command-line tool to inspect and diagnose SQL workload performance before production deployment.
Star history
How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.
