ra-h-os/QDRANT-DEPLOYMENT.md

Qdrant Deployment

Qdrant is the optional vector-search sidecar for environments where sqlite-vec is unavailable or unreliable.

Use Qdrant for:

• Alpine/musl Docker images
• Windows ARM64
• Linux ARM64 setups where sqlite-vec has not been validated
• any setup where sqlite-vec cannot be installed or loaded reliably

Qdrant is not required for local embeddings. The embedding model creates vectors; the vector backend stores and searches them.

Start Qdrant

docker compose up -d qdrant

Configure RA-H:

VECTOR_BACKEND=qdrant
QDRANT_URL=http://localhost:6333

Validate:

npm run doctor:local-ai

Rebuild

Switching from sqlite-vec to Qdrant, or changing embedding provider/model/dimensions, requires rebuilding vector indexes:

npm run rebuild:embeddings

If existing Qdrant collections have the wrong dimensions and you intentionally want to recreate them from SQLite source data:

QDRANT_RECREATE_COLLECTIONS=true npm run rebuild:embeddings

Nodes, edges, source text, chunks, and metadata remain in SQLite. Qdrant can be deleted and rebuilt from SQLite-derived source data.