services:
postgres:
image: pgvector/pgvector:pg16
environment:
POSTGRES_DB: berryrag
POSTGRES_USER: berryrag
POSTGRES_PASSWORD: berryrag_password
ports:
- "${POSTGRES_PORT:-5432}:5432"
volumes:
- postgres_data:/var/lib/postgresql/data
- ./init.sql:/docker-entrypoint-initdb.d/init.sql
healthcheck:
test: ["CMD-SHELL", "pg_isready -U berryrag -d berryrag"]
interval: 5s
timeout: 5s
retries: 5
app:
build: .
environment:
- DATABASE_URL=postgresql://berryrag:berryrag_password@postgres:5432/berryrag
- OPENAI_API_KEY=${OPENAI_API_KEY}
- EMBEDDING_PROVIDER=${EMBEDDING_PROVIDER:-auto}
- RAG_STORAGE_PATH=/app/storage
- SCRAPED_CONTENT_PATH=/app/scraped_content
- MIN_CONTENT_LENGTH=${MIN_CONTENT_LENGTH:-100}
- MAX_CONTENT_LENGTH=${MAX_CONTENT_LENGTH:-500000}
- CHUNK_SIZE=${CHUNK_SIZE:-500}
- CHUNK_OVERLAP=${CHUNK_OVERLAP:-50}
- DEFAULT_TOP_K=${DEFAULT_TOP_K:-5}
- SIMILARITY_THRESHOLD=${SIMILARITY_THRESHOLD:-0.1}
- MAX_CONTEXT_CHARS=${MAX_CONTEXT_CHARS:-4000}
- LOG_LEVEL=${LOG_LEVEL:-INFO}
# Playwright configuration
- PLAYWRIGHT_BROWSERS_PATH=/ms-playwright
- PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=0
volumes:
- ./storage:/app/storage
- ./scraped_content:/app/scraped_content
- playwright_browsers:/ms-playwright
depends_on:
postgres:
condition: service_healthy
ports:
- "${APP_PORT:-8000}:8000"
mcp-server:
build: .
command: ["node", "dist/vector_db_server.js"]
environment:
- DATABASE_URL=postgresql://berryrag:berryrag_password@postgres:5432/berryrag
- OPENAI_API_KEY=${OPENAI_API_KEY}
- EMBEDDING_PROVIDER=${EMBEDDING_PROVIDER:-auto}
- RAG_STORAGE_PATH=/app/storage
- SCRAPED_CONTENT_PATH=/app/scraped_content
- MIN_CONTENT_LENGTH=${MIN_CONTENT_LENGTH:-100}
- MAX_CONTENT_LENGTH=${MAX_CONTENT_LENGTH:-500000}
- CHUNK_SIZE=${CHUNK_SIZE:-500}
- CHUNK_OVERLAP=${CHUNK_OVERLAP:-50}
- DEFAULT_TOP_K=${DEFAULT_TOP_K:-5}
- SIMILARITY_THRESHOLD=${SIMILARITY_THRESHOLD:-0.1}
- MAX_CONTEXT_CHARS=${MAX_CONTEXT_CHARS:-4000}
- LOG_LEVEL=${LOG_LEVEL:-INFO}
volumes:
- ./storage:/app/storage
- ./scraped_content:/app/scraped_content
depends_on:
postgres:
condition: service_healthy
app:
condition: service_started
ports:
- "${BERRY_RAG_PORT:-3000}:3000"
playwright-service:
build: .
command: ["python", "src/playwright_integration.py", "process"]
environment:
- DATABASE_URL=postgresql://berryrag:berryrag_password@postgres:5432/berryrag
- OPENAI_API_KEY=${OPENAI_API_KEY}
- EMBEDDING_PROVIDER=${EMBEDDING_PROVIDER:-auto}
- RAG_STORAGE_PATH=/app/storage
- SCRAPED_CONTENT_PATH=/app/scraped_content
- MIN_CONTENT_LENGTH=${MIN_CONTENT_LENGTH:-100}
- MAX_CONTENT_LENGTH=${MAX_CONTENT_LENGTH:-500000}
- CHUNK_SIZE=${CHUNK_SIZE:-500}
- CHUNK_OVERLAP=${CHUNK_OVERLAP:-50}
- DEFAULT_TOP_K=${DEFAULT_TOP_K:-5}
- SIMILARITY_THRESHOLD=${SIMILARITY_THRESHOLD:-0.1}
- MAX_CONTEXT_CHARS=${MAX_CONTEXT_CHARS:-4000}
- LOG_LEVEL=${LOG_LEVEL:-INFO}
# Playwright configuration
- PLAYWRIGHT_BROWSERS_PATH=/ms-playwright
- PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=0
volumes:
- ./storage:/app/storage
- ./scraped_content:/app/scraped_content
- playwright_browsers:/ms-playwright
depends_on:
postgres:
condition: service_healthy
restart: "no" # Run once for processing
berry-exa-server:
build: .
command: ["python", "mcp_servers/berry_exa_server.py"]
environment:
- DATABASE_URL=postgresql://berryrag:berryrag_password@postgres:5432/berryrag
- OPENAI_API_KEY=${OPENAI_API_KEY}
- EMBEDDING_PROVIDER=${EMBEDDING_PROVIDER:-auto}
- RAG_STORAGE_PATH=/app/storage
- SCRAPED_CONTENT_PATH=/app/scraped_content
- MIN_CONTENT_LENGTH=${MIN_CONTENT_LENGTH:-100}
- MAX_CONTENT_LENGTH=${MAX_CONTENT_LENGTH:-500000}
- CHUNK_SIZE=${CHUNK_SIZE:-500}
- CHUNK_OVERLAP=${CHUNK_OVERLAP:-50}
- DEFAULT_TOP_K=${DEFAULT_TOP_K:-5}
- SIMILARITY_THRESHOLD=${SIMILARITY_THRESHOLD:-0.1}
- MAX_CONTEXT_CHARS=${MAX_CONTEXT_CHARS:-4000}
- LOG_LEVEL=${LOG_LEVEL:-INFO}
# Playwright configuration for web crawling
- PLAYWRIGHT_BROWSERS_PATH=/ms-playwright
- PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=0
volumes:
- ./storage:/app/storage
- ./scraped_content:/app/scraped_content
- playwright_browsers:/ms-playwright
depends_on:
postgres:
condition: service_healthy
app:
condition: service_started
ports:
- "${BERRY_EXA_PORT:-3001}:3001"
stdin_open: true
tty: true
volumes:
postgres_data:
playwright_browsers:
