Azure Functions Template

This repository acts as a template for Azure Functions development and deployment.

Prerequisites

Azure CLI
Azure Functions Core Tool
Azure Function App (for further detail on how to set it up, see the Deployment section below)

Description

The function_builder.bat file is a batch file designed to automate the setup of a Python-based Azure Function within a virtual environment. To activate it, just put the batch file inside the root directory and execute it.

Here are the main steps of the process:

Creating a Virtual Environment: The script creates a Python virtual environment in the current directory. This isolates the Python environment for the Azure Function, preventing any conflicts with other Python projects on the same system.
User Input for Function Name: The script prompts the user to enter a name for the Azure Function.
Creating the Azure Function: The script creates a new directory named “Azure Functions”, navigates into it, and then creates a new Azure Function with the provided name. The function is created with the Python worker runtime, using the v2 model, and the “Http Trigger” template for demo purposes.
Updating the Requirements File: The script navigates into the newly created Azure Function’s directory and overwrites the requirements.txt file with a new set of dependencies.

Next Steps (Local)

Activate the Virtual Environment

In order to start a virtual environment it is important to install the python module VENV.

With the following command you can create a virtual env in the current directory:

python -m venv .venv

To activate the virtual env you have to run one of the activation applications inside the terminal. For PowerShell use the following command:

PS C:\> <venv>\Scripts\Activate.ps1

Install Dependencies and Run Locally

# Install the dependencies inside the function directory
pip install -r ./requirements.txt

# Start the local azure function
func host start

If the function does not return an exception you should see the localhost:// paths that are now active.

Next Steps (Deployment)

# Make sure that you login with the azure CLI
az login

# Create a resource group
az group create --name <RESOURCEGROUPNAME> --location <LOCATION>

# Create a storage account
az storage account create --name <STORAGEACCOUNTNAME> --location <LOCATION> --resource-group <RESOURCEGROUPNAME> --sku Standard_LRS

# Create a functionapp
az functionapp create --resource-group <RESOURCEGROUPNAME> --consumption-plan-location <LOCATION> --runtime <RUNTIME> --functions-version 3 --name <FUNCTIONAPPNAME> --storage-account <STORAGEACCOUNTNAME> --os-type Linux

# To download remote application settings
func azure functionapp fetch-app-settings <FUNCTIONAPPNAME>

# To deploy your created function
func azure functionapp publish <FUNCTIONAPPNAME>

# To deploy your created function inside a specific deployment slot
func azure functionapp publish <FUNCTIONAPPNAME> --slot <SLOTNAME>

Deployment with Github Actions

In order to use Github Actions for automatic deployment you have to adjust some settings. First make sure that you save the publish profile from your function:

Ensure that the SCM Auth is enabled:
In your Repo, go to [SETTINGS > SECRETS AND VARIABLES > ACTIONS] and store the publish profile from your azure function.
Go to the workflow file (./.github/workflows/azure-functions.yml) and adjust the commented parts (the ones that start with “set this to your…”) to fit your own situation

Now when your code is ready for production, merge your code to the “azure-production” branch and the github action will deploy.

Warning: If you enter the github action like the following: Azure/functions-action@v1 it does not always use the latest release. As explained in the consumption plans section below, for flex consumption it is important that you at least use the 1.5.2 version: Azure/functions-action@v1.5.2

Consumption Plans

If you are working with large files and want the function to handle them, the default consumption with its default RAM is probably not enough. With this in mind, Microsoft created another consumption plan called Flex consumption. This consumption plan has double the RAM and other features (including virtual network support).

The downside though: some usual helpful features and properties are not (or not yet) supported. For further information regarding deprecations please click here.

This repository contains an exemplary workflow file (./.github/workflows/azure-functions.yml) that is set up for the flex consumption plan. If you want to use it, make sure that you remove the .example at the end of the yml extension and replace the current workflow.

Keep in mind that when you are using the flex consumption plan, setting up a connectionstring and app settings for the function app contain specific presets (for example: AzureWebJobs….)