Building a Milvus Cluster Based on JuiceFS
Jun 15, 2021Changjian Gao and Jingjing Jia

Summary
This post details the integration of Milvus with JuiceFS, a cloud-native distributed POSIX file system. It explains how to configure Milvus to use JuiceFS for shared file storage, including setting up MySQL for metadata, creating a JuiceFS file system on object storage (e.g., Azure Blob Storage) and Redis for metadata, and configuring Milvus nodes (read/write roles) and Mishards. Performance benchmarks are presented, highlighting JuiceFS's advantages over EFS in IOPS and throughput, and demonstrating near-instantaneous data loading from disk to memory.
More from the AI Data Storage Engine trail
- Milvus Snapshots: Point-in-Time Collection Views Without Copying Data
- Announcing Milvus 3.0: Lake-Native Vector Search and a More Powerful Retrieval Engine
- How to Recover Milvus Data After etcd Corruption Without a Backup
- Open Tag: An Open-Source Claude Tag for Claude Code and Codex
- How Force Merge Compaction Nearly Doubled Milvus Search QPS