AliSQL: Alibaba's open-source MySQL with vector and DuckDB engines
Alibaba has open-sourced AliSQL, a specialized MySQL fork that integrates DuckDB for analytical processing and an optimized HNSW algorithm for high-performance vector search. This release signifies a move towards enhancing MySQL's capabilities for Hybrid Transactional/Analytical Processing (HTAP) and AI-driven applications, allowing users to leverage advanced data functionalities directly within their familiar MySQL environment. The Hacker News discussion dives into the implications of this integration, comparing it to existing solutions and debating the true nature of its HTAP claims.
The Lowdown
AliSQL, Alibaba Group's production-hardened fork of MySQL, has been open-sourced, bringing advanced analytical and AI capabilities directly into the world's most popular open-source relational database.
- AliSQL is based on MySQL 8.0.44 (LTS), inheriting its core functionalities while adding Alibaba's optimizations.
- It integrates DuckDB as a native storage engine, enabling lightweight analytical queries directly within the MySQL ecosystem.
- A significant feature is its native support for enterprise-grade vector processing, utilizing an optimized HNSW algorithm for Approximate Nearest Neighbor (ANN) search across up to 16,383 dimensions, facilitating AI applications like semantic search.
- The roadmap hints at further optimizations for DDL operations, Recovery Time Objective (RTO), and replication, aiming for improved performance and stability.
- AliSQL is licensed under GPL-2.0, aligning with MySQL's open-source ethos.
This initiative by Alibaba extends MySQL's utility beyond traditional OLTP, positioning it as a more versatile database for mixed workloads and emerging AI-centric use cases.
The Gossip
HTAP Horizons and Database Integration
The discussion extensively covers AliSQL's approach to Hybrid Transactional/Analytical Processing (HTAP) by integrating DuckDB and vector search. Some commenters view this as a significant advancement, simplifying data pipelines and bringing analytical power to MySQL, drawing parallels to systems like SAP HANA. However, skepticism is also voiced regarding whether it offers true transactional consistency between the different engines or if it merely repackages existing integration patterns without providing unified HTAP guarantees. The practical necessity and desirability of transactional consistency for analytical and vector use cases also sparked debate.
MySQL's New Arsenal and Competitor Comparisons
Commenters ponder Alibaba's decision to extend MySQL rather than utilizing established extension mechanisms found in other databases, such as PostgreSQL's Foreign Data Wrappers (FDW) or its `pg_duckdb` extension. The conversation frequently compares AliSQL's features to those offered by other databases and technologies, including MariaDB's columnar engine and vector storage, ClickHouse, TiDB, and Timescale. Users are keen to understand if AliSQL's direct integration provides unique advantages or if it's simply a different implementation path to similar analytical and vector functionalities.
The Curious Case of the Commit Chronicle
A notable point of discussion was the project's GitHub commit history, which appeared unusually sparse, showing a few scattered commits over several years, with a 'first commit' curiously dated in the future (2026). This anomaly prompted speculation regarding the project's development and its transition to open source. A common explanation offered is that such histories are typical for internally developed projects that are later open-sourced, where the full, potentially messy or proprietary internal version control history is intentionally not made public.