Robot Ebert

A Conversational Movie Recommendation App

Access the Hosted (GCP Cloud Run) Streamlit Front-End Application

Robot Ebert

Run the Application Locally

Run the FastAPI Backend (Reload on Save)

cd src && PYTHONPATH=$PWD RELOAD=true python backend/app/main.py

Run the Streamlit Frontend (Reload on Save)

cd src && streamlit run frontend/app/main.py

Run the Application via Docker Desktop

Run the FastAPI Backend

cd src && docker build -f backend/Dockerfile -t robot-ebert-fastapi . && cd -
docker run -p 8080:8080 --env-file .env robot-ebert-fastapi

Run the Streamlit Frontend

cd src && docker build -f frontend/Dockerfile -t robot-ebert-streamlit . && cd -
docker run -p 8501:8501 --env-file .env robot-ebert-streamlit

Deploy the Application to GCP Cloud Run

Set Global Environment Variables

export PROJECT_ID="robot-ebert"
export LOCATION="us-west1"
export REPO_NAME="robot-ebert"
export REPO_DESCRIPTION="container images for the Robot Ebert movie recommender application"

Create the Repo in GCP Artifact Registry

gcloud auth login
gcloud artifacts repositories create ${REPO_NAME} --repository-format=docker --location=${LOCATION} --description=${REPO_DESCRIPTION}
gcloud artifacts repositories list
gcloud auth configure-docker ${LOCATION}-docker.pkg.dev

Update the FastAPI Backend Service

Set Environment Variables

export IMAGE_NAME="robot-ebert-fastapi"
export SERVICE_NAME="robot-ebert-fastapi"

Build and Push the FastAPI Image to Artifact Registry

cd src && docker build --platform linux/amd64 -f backend/Dockerfile -t ${IMAGE_NAME} . && cd -
docker tag ${IMAGE_NAME}:latest ${LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${IMAGE_NAME}:latest
docker push ${LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${IMAGE_NAME}:latest

Update the FastAPI Backend Cloud Run Service to Use the New Image

gcloud run deploy ${SERVICE_NAME} --image ${LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${IMAGE_NAME}:latest --platform managed --region $LOCATION

Update the Streamlit Frontend Service

Set Environment Variables

export IMAGE_NAME="robot-ebert-streamlit"
export SERVICE_NAME="robot-ebert-streamlit"

Build and Push the Streamlit Image to Artifact Registry

cd src && docker build --platform linux/amd64 -f frontend/Dockerfile -t ${IMAGE_NAME} . && cd -
docker tag ${IMAGE_NAME}:latest ${LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${IMAGE_NAME}:latest
docker push ${LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${IMAGE_NAME}:latest

Update the Frontend Cloud Run Service to Use the New Image

gcloud run deploy ${SERVICE_NAME} --image ${LOCATION}-docker.pkg.dev/${PROJECT_ID}/${REPO_NAME}/${IMAGE_NAME}:latest --platform managed --region $LOCATION

Connect to the Application Database in GCP CloudSQL

Apply the SQLAlchemy DDL to the Database to Create the Tables

gcloud auth application-default login
cd app && python database.py && cd -

Connect to the Database Locally using pgAdmin

• Whitelist Your Client IP: CloudSQL > Instances > ${INSTANCE_NAME} > Networking > Add a Network > ${CLIENT_PUBLIC_IP}
• Add a New Server in pgAdmin: Servers > Register > Server:
  • Host: ${SERVER_PUBLIC_IP}
  • User: postgres
  • Pass: POSTGRES_PASSWORD (GCP Secrets Manager)
  • Port: 5432