demo-streamlit-azure

• Demo repo to deploy Streamlit private repo to Azure service.
• Note: This guideline is written on August 2023. Any changing or deprecation might be happened in the future.

Deployment Guideline

Step 1: Basics Section

1. Search App Serives and click it. Then click Create and choose Web App.
2. Choose existed Resource Group (rg_invoke)
3. Write app name. Please ensure the name is descriptive as possible. ie. analytics-dashboard-web-app
4. For Publish, choose Code (if you have container, choose Container)
5. Choose runtime stack (Python). Please ensure the Python version is similar to the Python version that you used on local development.
6. Select region (Southeast Asia)
7. Select operating system (Linux is recommended)
8. Create Linux-plan name. Please ensure the name is descriptive as possible. ie linux-analytics
9. Choose pricing plan for compute resources. Refer experience developer/lead to get some advice on the choice.
10. Ignore Zone redundancy.
11. Move to Step 2.

Step 2: Deployment

1. On the same page, click Deployment tab.
2. Click Enable for Continuous deployment.
3. On Github account, link to your Github account.
4. Choose organization (INVOKE-Analytics)
5. Select project repository name.
6. Select branch to deploy.
7. Click Review + create. Please ensure all the description is correct following to your deployment plan.
8. Click Create.
9. Move to Step 3.

Step 3: Configuration

1. Click your web-app name in App Service.
2. Click Configuration (Under settings)
3. Click General settings
4. Ensure the stack is Python, major version and minor version is correct.
5. Insert the following line for startup command. The main.py is referred to the top script to deploy.

python -m streamlit run main.py --server.port 8000 --server.address 0.0.0.0

6. Click Save.
7. Move to step 4.

Step 4: Github Actions

1. Open your project repo on Github.
2. You will see there is .github/workflows directory created by Azure for the CI/CD.
3. Click Actions.
4. A workflow should be running.
5. Wait until the process is finished.
6. If the workflow failed, rerun the workflow.
7. If the workflow succeed, click the link return by the workflow.
8. You will be redirected to your Streamlit app.
9. Deployment finished.
10. Congratulation you have deployed your app on Azure!