TideSQL: New Space-Efficient Storage Engine for MariaDB Offers ACID Transactions and Lock-Free Concurrency
Key Takeaways
- ▸TideSQL enables lock-free ACID transactions through MVCC, allowing concurrent reads and writes without blocking
- ▸The engine provides space-efficient storage optimization and is write-optimized for high-throughput workloads
- ▸Seamless migration from InnoDB requires only changing the ENGINE clause, with no application code modifications needed
Summary
TideSQL is a pluggable storage engine for MariaDB built on top of TidesDB, designed to provide space-efficient and write-optimized database storage. The engine supports comprehensive features including ACID transactions through multi-version concurrency control (MVCC), secondary indexes, auto-increment columns, TTL-based expiration, encryption at rest, and online DDL operations. Users can seamlessly switch from InnoDB to TidesDB by simply changing the ENGINE clause in their SQL statements, making migration straightforward for existing MariaDB deployments.
The architecture separates each table into its own main column family with dedicated column families for secondary indexes, enabling independent LSM-trees with separate memtables, SSTables, and compaction schedules. TideSQL is distributed as a pluggable shared library that can be loaded at server startup or dynamically, and the project includes a cross-platform installation script that automates the entire setup process, handling dependencies, MariaDB source cloning, compilation, and configuration.
- Each table and secondary index operates with independent LSM-trees, enabling optimized compaction and flush schedules per index
- Installation is simplified through an automated cross-platform setup script that handles all dependencies and configuration
Editorial Opinion
TideSQL represents a meaningful alternative to traditional MySQL/MariaDB storage engines by combining modern LSM-tree architecture with comprehensive feature parity. The lock-free concurrency model and space-efficient design address real pain points for write-heavy workloads, while the simple migration path from InnoDB reduces barriers to adoption. However, the success of this engine will ultimately depend on real-world performance benchmarks, community adoption, and how well it handles production workload patterns that differ from traditional OLTP scenarios.
