update with pgvector setup and tested docker

.env

@@ -1,5 +0,0 @@
-OPENAI_API_KEY = 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
-PG_VECTOR_HOST = db  # Use the service name 'db' instead of localhost
-PG_VECTOR_USER = myuser
-PG_VECTOR_PASSWORD = mypassword
-PGDATABASE = mydb

Dockerfile

@@ -22,12 +22,6 @@ RUN pip install --no-cache-dir -r requirements.txt
 # Expose the port Streamlit runs on
 EXPOSE 8501
 
-# # Set the default command to run the app
-# CMD ["streamlit", "run", "app.py", "--server.port=8501", "--server.address=0.0.0.0"]
-# Set the default command to run the app
-# CMD ["streamlit", "run", "app.py", "--server.port=${PORT:-8501}", "--server.address=0.0.0.0", "--server.enableCORS=false", "--server.enableXsrfProtection=false"]
-
-# ENTRYPOINT ["streamlit", "run", "app.py", "--server.port=8501", "--server.address=0.0.0.0"]
 
 ENTRYPOINT ["bash", "-c", "streamlit run app.py --server.port=${PORT:-8501} --server.address=0.0.0.0"]
 

database.py

@@ -6,5 +6,6 @@ host = os.getenv("PG_VECTOR_HOST")
 user = os.getenv("PG_VECTOR_USER")
 password = os.getenv("PG_VECTOR_PASSWORD")
 COLLECTION_NAME = os.getenv("PGDATABASE")
-CONNECTION_STRING = "postgresql+psycopg://langchain:langchain@localhost:6024/langchain"
-#CONNECTION_STRING = f"postgresql+psycopg://{user}:{password}@{host}:6042/{COLLECTION_NAME}"
+CONNECTION_STRING = f"postgresql+psycopg://{user}:{password}@{host}:5432/{COLLECTION_NAME}"
+
+

docker-compose.yml

@@ -9,11 +9,14 @@ services:
     depends_on:
       redis-stack:
         condition: service_healthy  # Wait for Redis to be ready
+      db:
+        condition: service_healthy  # Wait for PostgreSQL to be ready
     env_file:
       - .env  # Load environment variables from .env file
     environment:
       - REDIS_HOST=redis-stack
       - REDIS_PORT=6379
+      - DATABASE_URL=postgresql+psycopg://${PG_VECTOR_USER}:${PG_VECTOR_PASSWORD}@${PG_VECTOR_HOST}:5432/${PGDATABASE}
     networks:
       - app_network
 
@@ -29,6 +32,33 @@ services:
       timeout: 3s
       retries: 5
 
+  db:
+    image: pgvector/pgvector:pg16  # Use PostgreSQL with pgvector
+    container_name: pgvector_db
+    restart: always
+    env_file:
+      - .env  # Load credentials from .env file
+    environment:
+      POSTGRES_USER: ${PG_VECTOR_USER}
+      POSTGRES_PASSWORD: ${PG_VECTOR_PASSWORD}
+      POSTGRES_DB: ${PGDATABASE}
+    ports:
+      - "6024:5432"
+    volumes:
+      - pgvector_data:/var/lib/postgresql/data  # Persist database
+      - ./init-scripts:/docker-entrypoint-initdb.d  # Auto-run SQL scripts
+    networks:
+      - app_network
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U ${PG_VECTOR_USER} -d ${PGDATABASE}"]
+      interval: 10s
+      timeout: 5s
+      retries: 5
+
+volumes:
+  pgvector_data:
+    driver: local
+
 networks:
   app_network:
     driver: bridge