What is Milvus
Milvus is an open source vector database that supports addition, deletion, modification and near-real-time query of TB-level vectors. It is highly flexible, stable, reliable, and high-speed query. Milvus integrates widely used vector index libraries such as Faiss, NMSLIB, and Annoy, and provides a set of simple and intuitive APIs, allowing you to choose different index types for different scenarios. In addition, Milvus can also filter scalar data, further improving recall and enhancing search flexibility.
Milvus server adopts master-slave architecture (Client-server model).
- On the server side, Milvus consists of Milvus Core and Meta Store:
-
- Milvus Core stores and manages vector and scalar data.
- Meta Store stores and manages metadata in SQLite and MySQL for testing and production respectively.
- On the client side, Milvus also provides SDK and RESTful API based on Python, Java, Go, and C++.
Milvus was released under the Apache 2 License and was officially open sourced in October 2019. It is an incubation project of the LF AI & DATA Foundation . The source code of Milvus is hosted on Github .