Skip to content

Materials for my session at PyCon 2023 - Accelerate your workflow from local Python prototype to the cloud

License

Notifications You must be signed in to change notification settings

savannahostrowski/pycon-2023

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Accelerate your workflow from local Python prototype to the cloud (PyCon US 2023)

You can watch a recording of this session on YouTube.

This repo hosts materials for my session at PyCon US 2023.

  • Have you ever struggled taking a Python application from your local machine to the cloud?
  • Had a hard time figuring out what infrastructure you need or how to configure it for your app?
  • Spent too much time researching how to set up your local development environment for cloud development?

Then, this talk/repo is for you!

Learn how to use real-world cloud development application templates via CLI to go from local development environment to the cloud. Scaffold your application, provision resources, deploy code, monitor your application health, and set up a CI/CD pipeline, all in a couple of steps and just a few minutes.

This talk is 1 part talk (slides) and 1 part workshop/DIY!

Workshop/DIY Prerequisites

If you are attending this workshop in person at PyCon, you should redeem your Azure credits at microsoftazurepass.com and follow these Azure Pass redemption instructions before starting!

For the core azd experience, you'll want to:

For the template, you'll want to install:

Note: if you choose another template, then you will need to install the dependencies as they are listed in that template's README. They may or may not be the same as listed above.

In VS Code, you'll want to install:

The TL;DR (aka the quick quickstart)

Please install the Azure Developer CLI and template dependencies(instructions above) before tackling the quickstart. You can also try this same quickstart with any template found on aka.ms/awesome-azd. If you choose another template, just replace the template parameter in the init command with the GitHub URL to the template repo.

Sometimes, you want to just get right into it and read about all the specifics later! That's cool. Run through these commands to authenticate on Azure, initialize using a template, provision infrastructure and deploy code on Azure.

  1. azd auth login
  2. azd init --template Azure-Samples/todo-python-mongo-aca (in an empty directory)
  3. azd up

As a bonus, you can also run these commands to see monitoring dashboards and set up a CI/CD pipeline (if the template has dashboard infrastructure and a GitHub Actions or Azure Pipelines definition):

  1. azd monitor
  2. azd pipeline config

Quickstart tutorial with the details

Please install the Azure Developer CLI and template dependencies(instructions above) before tackling the quickstart. You can also try this same quickstart with any template found on aka.ms/awesome-azd.

Authenticate!

Run azd auth login in your terminal.

This will open a browser for you to authenticate on Azure.

Let's grab an application template!

For this workshop/example, we're going to get https://github.com/Azure-Samples/todo-python-mongo-aca up and running on Azure in just a couple of minutes. This template uses React for the frontend, Python with FastAPI for the backend, Azure Cosmos DB for MongoDB for storage, and Azure Monitor for monitoring and logging. The template includes sample application code (a ToDo application) included as a proof-of-concept, infrastructure as code, and more!

Azure Developer CLI-compatible repos/templates are designed to be extensible and easy to customize. The idea is that you can find a template that matches your application's architecture (e.g. frontend, backend, database) and use the template as a foundational blueprint for getting your application up on Azure.

Alright, let's grab the application template code.

  1. Make a new directory - mkdir pycon-azd-todo
  2. Change into that directory - cd pycon-azd-todo
  3. Run azd init --template Azure-Samples/todo-python-mongo-aca.

After running this azd init command, you'll be prompted for an environment name. The environment name is not your virtual environment. You should name it something that can help you identify the project and stage i.e. todo-dev etc.

Finally, let's open the project in VS Code by typing code ..

Provision infrastructure and deploy application code

Now, let's run the azd up command. This command is a special, workflow-type command which will:

  • Package your application code - in this case, we are going to build container images using Docker
  • Provision infrastructure - as defined in infra/
  • Deploy code on Azure - to Azure Container Apps

This step will take a couple of minutes. With this command, azd isn't just getting your app hosted on Azure. It's setting up roles and permissions, Key Vault for storing secrets, wiring up our database, and more!

Options for running up in VS Code:

  1. azd up in the terminal
  2. Right-click on the azure.yaml in the root of the template and select up from the context menu
  3. Open the command palette (ctrl+shift+p on Windows or cmd+shift+p on Mac), type "azd up" and select "Azure Developer: up"

After running this command, you'll be prompted for:

  • An Azure subscription to use
  • A location (eastus, westus, etc.) where your resources will be deployed

At the end of this step you'll be able to see two endpoints for the frontend and backend of the app, all up on Azure.

Tip: While azd up is running, take a look at what's included in the template! You'll notice that the template has:

  • Real application code (src/) - provided as a proof-of-concept
  • A .vscode/ - includes a launch.json file to define debug configurations and a tasks.json file to define the configurations to start the web or API server for local development
  • A .devcontainer/ - to define a Devcontainer/Codespaces configuration
  • An infra/ - hosts a collection of infrastructure as code assets written either in Bicep or Terraform to provision the right resources, deploy code on Azure and set up application monitoring
  • A .github/ and .azdo/ - contains a GitHub Actions workflow and Azure Pipeline definition respectively to test your application against real Azure resources on every commit
  • A azure.yaml - an azd specific file that defines application metadata (e.g. languages, hosts, paths to code needed to build the app)

Exploring resources using the Azure Tools extension pack for VS Code

We can also install the Azure Tools extension pack for VS Code which will we can also see what resources and services were created after running azd up in an application view.

  1. Click on the Azure icon in the activity bar.

  2. In the workspace view, you'll see that we have a visual representation of our azure.yaml.

  3. Right-click on the web node under Services and notice that there are contextually relevant azd commands like azd deploy (which would allow us to deploy our code on Azure for the frontend).

  4. We could even click on the "cloud down" icon to run azd down and destroy all the infrastructure and deployments related to this application.

Best practices - monitoring and CI/CD

We all know that it's not enough to just get your application up on Azure. We want to set ourselves up for success in the future. Let's talk about monitoring and setting up CI/CD.

Monitor your app

As part of azd provision (which was run as part of the azd up command), we created some application monitoring infrastucture. In the infra/main.bicep of https://github.com/Azure-Samples/todo-python-mongo-aca, you'll find a monitoring module which created all the monitoring infrastructure and a dashboard for our application. By running the following command, you will be brought to that dashboard and see usage, performance and reliability metrics for your application.

Options for running monitor in VS Code:

  1. azd monitor in the terminal
  2. Right-click on the azure.yaml in the root of the template and select monitor from the context menu. Select "Application Insights Overview Dashboard" from the list and then "OK".
  3. Open the command palette (ctrl+shift+p on Windows or cmd+shift+p on Mac), type "azd monitor" and select "Azure Developer: monitor".

This will bring you to a dashboard with performance, reliability and usage metrics for the deployed application.

Configure a CI/CD pipeline with GitHub Actions

Finally, let's talk about setting up CI/CD. Included in the template is a GitHub Action workflow file. To configure and enable the action, azd provides the pipeline config command/action which performs the following:

  • Creates and configures the Service Principal on the Azure subscription
  • If there is not a git repository configured, provides the needed mechanics to create/configure your repository as well as commit the code.
  • Create a secure connection between Azure and your repository using GitHub secrets.
  • Upon checking in the workflow file, runs the action.

Again, you can either run azd pipeline config via:

  1. azd pipeline config in the terminal
  2. Right-click on the azure.yaml in the root of the template and select pipeline config from the context menu.
  3. Open the command palette (ctrl+shift+p on Windows or cmd+shift+p on Mac), type "azd pipeline config" and select "Azure Developer: pipeline config"

After this completes and your code is pushed to a GitHub repo, open up the Actions tab and see that a run of the pipeline is kicked off!

Congrats! You deployed an application on Azure!

What's next? You can:

  • Try out the same workflow above with another Python template on the Azure Developer CLI template gallery.
  • Try adding your own application code to a template
  • Try building your own template
    • If you create your own template/azd-compatible application, we'd love to add it to the gallery. Submit a PR!

Additional Resources

If you're interested in learning more, be sure to check out:

About

Materials for my session at PyCon 2023 - Accelerate your workflow from local Python prototype to the cloud

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published