apache/shardingsphere

This system is a comprehensive database middleware and agent framework that provides database connectivity, protocol handling, and runtime instrumentation capabilities. It supports a wide array of database dialects including MySQL, Oracle, PostgreSQL, and others, while offering features like read-write splitting, sharding, encryption, and data masking. Additionally, it includes a Java agent for collecting metrics, tracing, and logging via OpenTelemetry and Prometheus, alongside a SQL parser and execution engine for distributed database operations.

This period focused on establishing a robust, plugin-based architecture for the ShardingSphere Agent, introducing a standardized lifecycle management system and YAML configuration support. Simultaneously, the core kernel underwent significant refactoring to support new features like global clock coordination, SQL federation, and enhanced transaction management, while also improving the proxy and standalone mode implementations.

173 changes

The period focused on expanding database connectivity by implementing full connector and protocol support for Firebird, Hive, MariaDB, and SQL Server, alongside significant refactoring of the database connector core to standardize metadata loading and JDBC URL parsing. Concurrently, the codebase introduced new DistSQL handlers and RAL statements for managing data pipelines, CDC streaming, and migration jobs, while also establishing a more modular architecture for SQL binding, routing, and rewriting across multiple database dialects.

166 changes

This period focused on establishing the Model Context Protocol (MCP) integration, introducing core APIs, bootstrap mechanisms, and feature-specific handlers for encryption, masking, broadcast, readwrite-splitting, and sharding. Concurrently, the codebase saw a significant expansion of unit and end-to-end test coverage for these new features, alongside improvements to identifier case policies and Firebird exception mapping.

36 changes

apache/shardingsphere was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point the surveyor at a repository you know and see whether you agree with it.