Deploy Milvus

Milvus vector database v3.0.6 with Attu web UI, auth on, data on a volume

Deploy Milvus

/var/lib/milvus

/data

Deploy and Host Milvus on Railway

Milvus is an open-source vector database (Apache-2.0) for similarity search over embeddings. It stores dense and sparse vectors next to scalar fields, builds ANN indexes such as HNSW and IVF, and answers nearest-neighbour, hybrid and full-text queries through gRPC SDKs (Python, Node.js, Go, Java) and a REST API.

About Hosting Milvus

This template runs Milvus v3.0.6 in standalone mode. Etcd is embedded and storage is local, the same single-container setup as upstream's standalone_embed.sh, so metadata, the write-ahead log and segments all live on one volume. Authentication is always on, and the root password is generated per deploy. The template also adds Attu v3.0.2, the official Milvus web UI. Its admin password is generated too, and its connection to Milvus over the private network is set up for you. Both services build from a public GitHub repo that pins each image by version and digest, so Railway can tell you when we publish an update.

Common Use Cases

  • Retrieval-augmented generation (RAG): store document chunks as embeddings and fetch the best context for an LLM
  • Semantic, hybrid and full-text search over products, documents, images or code
  • Recommendations, deduplication and long-term memory for AI agents

Dependencies for Milvus Hosting

  • Milvus milvusdb/milvus:v3.0.6 (standalone, embedded etcd, local storage) with a volume at /var/lib/milvus
  • Attu zilliz/attu:v3.0.2 (web UI) with a volume at /data

Deployment Dependencies

Implementation Details

After deploy:

  • Web UI: open the Attu service's domain and sign in as admin with the ATTU_ADMIN_PASSWORD variable from the Attu service. The Milvus connection is already there.
  • From another Railway service in the same project: use the URI http://${{Milvus.RAILWAY_PRIVATE_DOMAIN}}:19530 and the token root:${{Milvus.COMMON_SECURITY_DEFAULTROOTPASSWORD}}. This works for pymilvus and the other gRPC SDKs as well as for REST.
  • From outside Railway: send REST v2 calls to the Milvus service's domain, for example POST https://YOUR-MILVUS-DOMAIN/v2/vectordb/collections/list with the header Authorization: Bearer root:YOUR-PASSWORD. Railway's HTTPS domains don't carry native gRPC. For gRPC SDKs from outside, add a TCP proxy to port 19530 yourself, but note that its traffic is not encrypted.
  • Passwords: the generated passwords are used on the first boot only. To change the root password, use Attu or the Milvus API, then update MILVUS_PASSWORD on Attu.
  • Updates: Railway shows an update when we bump the pinned versions in the repo. Read CHANGELOG.md first, and back up the Milvus volume before a major upgrade, because Milvus upgrades are one-way.

Why Deploy Milvus on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying Milvus on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.


Template Content

More templates in this category

View Template
Chat Chat
Chat Chat, your own unified chat and search to AI platform.

okisdev
117
View Template
Weaviate
Vector database with hybrid search and an API key-protected API

GitHub Locker
1
View Template
Möbius Personal AI OS
Self-hosted AI agent that builds and runs apps for you, with memory.

bento
1