TidesDB
A high-performance, embeddable LSM-tree storage engine library in C with ACID transactions, tiered storage, and cross-platform portability.
What is TidesDB?
TidesDB is a high-performance, embeddable storage engine library written in C, based on a log-structured merge-tree (LSM-tree). It provides ACID transactions, multi-column family support, and adaptive compaction, designed for flash and RAM optimization with optional tiered object storage for infinite scalability. It serves as a foundational component for building custom databases or as a standalone key-value/column store.
Database engineers and developers building custom database systems, embedded applications requiring transactional storage, or projects needing a portable, high-performance key-value store with advanced features like MVCC and tiered storage.
Developers choose TidesDB for its combination of embeddability, cross-platform portability, and rich feature set including ACID transactions with multiple isolation levels, adaptive compaction strategies, and optional S3-compatible object storage, all in a clean C API optimized for modern hardware.
Overview
High-performance adaptive, durable, portable, transactional embeddable storage engine with optional tiered object storage for infinite scale. Designed for flash and RAM optimization.
Use Cases
Best For
- Building custom transactional databases on top of an LSM-tree storage engine
- Embedded applications requiring ACID-compliant key-value storage with MVCC
- Systems needing cross-platform portable database files across diverse architectures
- Implementing tiered storage with S3-compatible object storage and local caching
- Optimizing write amplification and space usage with adaptive compaction strategies
- Developing high-performance storage solutions with configurable column families and isolation levels
Not Ideal For
- Applications requiring a full SQL database with query language and schema management
- Distributed systems needing built-in replication and sharding without custom implementation
- Teams seeking a managed database service with automatic scaling and minimal operational overhead
Pros & Cons
Pros
Supports five isolation levels including SERIALIZABLE using Serializable Snapshot Isolation (SSI), with savepoints and read-your-own-writes semantics, providing robust transactional guarantees for embedded applications.
Offers three compaction modes with Dynamic Capacity Adaptation (DCA) to dynamically scale level capacities and minimize write amplification, optimizing for both space and performance based on data size.
Runs on Linux, macOS, Windows, BSD, and Solaris/Illumos across x86, ARM, RISC-V, and PowerPC architectures, with little-endian file serialization ensuring database files work seamlessly across platforms without conversion.
Includes S3-compatible tiered object storage with local file caching and node-failure recovery, enabling infinite scalability for large datasets without sacrificing local performance.
Cons
Lacks higher-level features like SQL queries, a query engine, or schema enforcement, requiring developers to build these components on top if needed, which adds development complexity.
With numerous settings for compaction, caching, compression, and column families, tuning TidesDB for optimal performance can be overwhelming and prone to misconfiguration, especially for those new to LSM-trees.
As a C library, it may lack official or mature bindings for popular languages like Python, Java, or Rust, forcing teams to rely on community contributions or develop custom integrations.
Frequently Asked Questions
Related Projects
LevelDBLevelDB is a fast key-value storage library written at Google that provides an ordered mapping from string keys to string values.
RocksDBA library that provides an embeddable, persistent key-value store for fast storage.
HiredisMinimalistic C client for Redis >= 1.2
InfinityThe AI-native database built for LLM applications, providing incredibly fast hybrid search of dense vector, sparse vector, tensor (multi-vector), and full-text.
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.