bloomkv 1.3.1

A persistent key-value storage engine based on LSM-tree architecture.

BloomKV

Light weight key value storage engine optimized for write heavy workloads.

Overview

This project is an LSM-tree based key-value storage engine implemented in Python, optimized for write-heavy workloads.

It is a persistent key-value store featuring:

• A Write Ahead Log (WAL) for atomicity and durability.
• An in-memory Memtable using SortedDict.
• Persistent Sorted String Tables (SSTables).
• Asynchronous, leveled compaction handled by a background worker thread (introduced in v1.0.0).
• Metadata generation for created collections
• Bloom filter functionality to prevent unnecessary disk operations for keys that we know dont exist in the collection
• Storage engine is behind a FastAPI server and can be communicated with using REST APIs.
• Supports range queries from start_key to end_key
• Can be used as an importable library for programmatic integration
• Import csv files and store them as kv pairs in a collection

Getting Started

• Detailed documentation, usage and explanation of each component is available in the docs folder.

Contributing

Contributions, issues and feature requests are welcome! Contributing ideas

• Implement a proper testing suite for different components. Unit tests, Integration Tests
• Thread safety for multi threaded programs
• UI features along with CLI and library utilisation. and many more, ideas are always welcome!

LICENSE

This project is licensed under the MIT License. See the LICENSE file for details.