Deployment

Vectora is designed to be deployed flexibly, from local machines to cloud-based Kubernetes clusters.

The vectora-cloud binary is optimized for container environments, removing GUI dependencies and focusing on API and search engine performance.

Deployment Options

Vectora offers three main paths for deployment:

1. Docker: Ideal for isolated environments and quick testing.
2. Kubernetes: Recommended for large-scale production, with support for secrets and auto-scaling.
3. Native Binary: For local use or dedicated servers where containerization overhead is not desired.

Infrastructure Requirements

For a stable deployment, ensure you have:

• MongoDB Atlas: Recommended for vector and metadata persistence.
• API Keys: GOOGLE_API_KEY (Gemini) and VOYAGE_API_KEY (Voyage AI).
• Resources: Minimum 512MB RAM and 0.5 CPU per instance.

External Linking

Infrastructure & Orchestration