Google Cloud Function Scaffold

This repository provides a scaffold for building and developing Google Cloud Functions on your local machine. It is designed for building HTTP-triggered functions, and is recommended to be paried with something like Postman.

Dependencies

• Docker
• gcloud command-line tool
• GNU Make

Installation

To install, make sure all dependencies are first met, then run the following command to build the docker image and install required dependencies.

# Builds a python-flask image and installs all required dependencies as specified by your project
make build

Running Locally

To bring up your repository, use the make run command. This command simply echos docker-compose up, so feel free to use that instead.

# Starts a local python-flask container for serving your cloud function locally
make run
# To stop, use CTRL+C

This will bring up a development server on 0.0.0.0:8000, allowing you to preview and work with your function before deployment.

Available Make Commands

• make run - Will bring up the dev server in the current terminal
• make up - Will bring up the dev server in detached mode
• make down - Stops the dev server (even if running in detached mode)
• make stop - Stops the dev server (same as make down)
• make rebuild - Stops the dev server, rebuilds the image, and starts the server
• make build - Builds the dev server image
• make deploy - Deploys the current function to GCP

Adding PyPI Requirements

When requirements are added to the cloud_function/requirements.txt file, you will need to trigger a rebuild of the docker image to install these in the container. You can do so with the following command:

# Stops the currently running containers, rebuilds them, and starts them again
make rebuild

Deployment

To deploy to Google Cloud Functions, configure the settings.txt file with the following parameters (in order). Be sure there are no blank lines at the start of the file.

your-gcp-project-id
your_cloud_function_name

Once complete, you can run the make deploy command to deploy your cloud function to GCP.