Skip to content

armgit / ODS

Public Imported from GitHub
0

ODS V3 Pre-Release: Public testing and refinement ahead of the official V3 launch. Turn your PC, Mac, or Linux box into a private AI server.

3.5 KB
Raw Blame

Qdrant

Vector database for semantic search and RAG in ODS

Overview

Qdrant is a high-performance vector database that stores and searches embeddings for Retrieval-Augmented Generation (RAG) workflows. It enables semantic similarity search across your local documents, letting LLMs retrieve relevant context before answering questions.

Features

  • Vector storage: Persist high-dimensional embeddings with HNSW indexing for fast nearest-neighbor search
  • Collection management: Organize vectors into named collections with configurable distance metrics (cosine, dot product, Euclidean)
  • Filtered search: Combine semantic search with structured metadata filters
  • Snapshots: Create and restore point-in-time backups of your collections
  • REST + gRPC API: Full API access over HTTP and gRPC for high-throughput clients
  • Persistence: All data survives container restarts via bind-mounted storage

Configuration

Environment variables (set in .env):

Variable Default Description
QDRANT_PORT 6333 External HTTP REST API port
QDRANT_GRPC_PORT 6334 External gRPC API port

API Endpoints

Endpoint Method Description
GET / GET Health check / root info
GET /collections GET List all collections
PUT /collections/{name} PUT Create a collection
DELETE /collections/{name} DELETE Delete a collection
PUT /collections/{name}/points PUT Insert/update vectors
POST /collections/{name}/points/search POST Similarity search
GET /collections/{name}/points/{id} GET Get point by ID
GET /dashboard GET Built-in web dashboard

Full API docs are available at http://localhost:6333/dashboard when the service is running.

Architecture

┌──────────────┐     REST :6333    ┌──────────────┐
│  Open-WebUI  │──────────────────▶│    Qdrant    │
│  Perplexica  │   gRPC  :6334    │  (Vector DB) │
│  Your App    │──────────────────▶│              │
└──────────────┘                   └──────┬───────┘
                                          │
                                   ┌──────▼───────┐
                                   │ ./data/qdrant│
                                   │  (storage)   │
                                   └──────────────┘

Qdrant works alongside the embeddings service: the embeddings service converts text to vectors, and Qdrant stores and retrieves them.

Files

  • manifest.yaml — Service metadata (port, health endpoint, GPU backends)
  • compose.yaml — Container definition (image, volumes, ports, healthcheck)

Troubleshooting

Qdrant not starting:

docker compose ps ods-qdrant
docker compose logs ods-qdrant

Cannot connect on port 6333:

  • Check QDRANT_PORT in .env is not in use by another service
  • Verify the container is healthy: docker compose ps ods-qdrant

Data lost after restart:

  • Ensure ./data/qdrant directory exists and has correct permissions
  • Check volume mount in compose.yaml

Collection errors:

# List collections via REST API
curl http://localhost:6333/collections

# Check Qdrant version
curl http://localhost:6333/