Skip to content

containerization.md

Latest commit

 

History

History
84 lines (58 loc) · 2.16 KB

containerization.md

File metadata and controls

84 lines (58 loc) · 2.16 KB
Raw

Containerization Guide

This guide covers production-friendly container workflows for CORTEXA daemon runtime.

← Back to README

What you get

  • multi-stage Docker build
  • non-root runtime user
  • healthcheck on /health
  • optional Docker Compose setup with Qdrant

Deployment topology

flowchart LR
  CLIENT[Local client or service] --> DAEMON[CORTEXA daemon container]
  DAEMON --> DB[(SQLite volume)]
  DAEMON --> VECTOR[(Vector provider: memory or Qdrant)]
Loading
Profile Components Best for
Single container daemon + SQLite volume + memory vector provider local validation and lightweight demos
Compose stack daemon + SQLite volume + Qdrant team reproducibility and parity testing
Hardened runtime compose stack + TLS proxy + secrets manager production-like deployment posture

Build image

docker build -t cortexa:latest .

Run image (memory vector provider)

docker run --rm -p 4312:4312 -p 4321:4321 \
  -e CORTEXA_VECTOR_PROVIDER=memory \
  -e CORTEXA_DB_PATH=/var/lib/cortexa/cortexa.db \
  -v cortexa-data:/var/lib/cortexa \
  cortexa:latest

Run with Docker Compose (daemon + qdrant)

docker compose up --build

This starts:

  • cortexa daemon (4312, 4321)
  • qdrant vector store (6333)

Environment variables to tune

  • CORTEXA_DAEMON_PORT (default 4312)
  • CORTEXA_WS_PORT (default 4321)
  • CORTEXA_DB_PATH (default data/cortexa.db; in container use /var/lib/cortexa/cortexa.db)
  • CORTEXA_VECTOR_PROVIDER (memory, qdrant, chroma)
  • CORTEXA_VECTOR_URL (for qdrant/chroma backends)
  • CORTEXA_DAEMON_TOKEN (recommended for non-local access)

Quick health check

curl -s http://localhost:4312/health

Notes

  • For local-first development, pnpm run cortexa:daemon remains the fastest loop.
  • For team reproducibility and deployment parity, prefer Docker Compose with pinned image tags.