Build a function with buildpacks

This guide shows you how to use buildpacks with your function source code to create a container image. For example, use buildpacks to build a Cloud Run function that you want to deploy on Cloud Run.

There are two methods for building container images with buildpacks:

Build locally with the pack CLI to locally test your function and rapidly prototype changes before deployment.
Build remotely with Cloud Build. Building with Cloud Build is useful for functions that have a resource-intensive build processes and can also help protect your software supply chain.

Configure your project to build functions

To build functions with buildpacks:

Include the Functions Framework library.
Set the GOOGLE_FUNCTION_TARGET environment variable to the name of the function that you use as the entrypoint. You can do this by including a project.toml in the same folder as your source code. The project.toml file must have the following configuration:

[[build.env]]
    name = "GOOGLE_FUNCTION_TARGET"
    value =  "ENTRY_POINT"

Replace ENTRY_POINT with the function method.

For details about how to use environment variables with Cloud Run functions, see Configure Cloud Run functions services.

Local builds

Pack is a CLI tool maintained by the CNB project to support the use of buildpacks. Use the pack CLI to locally build your functions into a container image.

Before you begin

Install Docker Community Edition (CE) on your workstation. Docker is used by pack as an OCI image builder.
Install Pack CLI.
Install the Git source control tool to fetch the sample application from GitHub.

Build a function locally

You use the pack build command and specify the default builder --builder=gcr.io/buildpacks/builder to build your container images locally.

pack build --builder=gcr.io/buildpacks/builder IMAGE_NAME

Replace IMAGE_NAME with the name of your container image.

You can also customize your container image by extending the build and run images.

Build a sample function locally

The following examples demonstrate how to build a sample locally.

Clone the sample repository to your local machine:

git clone https://github.com/GoogleCloudPlatform/buildpack-samples.git

Change to the directory that contains the application sample code:

Go

cd buildpack-samples/sample-functions-framework-go

Java

cd buildpack-samples/sample-functions-framework-java-mvn

Node.js

cd buildpack-samples/sample-functions-framework-node

Python

cd buildpack-samples/sample-functions-framework-python

Ruby

cd buildpack-samples/sample-functions-framework-ruby

Use pack to build the sample function:

Go

pack build --builder=gcr.io/buildpacks/builder sample-functions-framework-go

Java

pack build --builder gcr.io/buildpacks/builder:v1 sample-functions-java-mvn

Node.js

pack build --builder=gcr.io/buildpacks/builder sample-functions-framework-node

Python

pack build --builder=gcr.io/buildpacks/builder sample-functions-framework-python

Ruby

pack build --builder=gcr.io/buildpacks/builder sample-functions-framework-ruby

Run the image using docker:

Go

docker run -p8080:8080 sample-functions-framework-go

Java

docker run -it -ePORT=8080 -p8080:8080 sample-functions-java-mvn

Node.js

docker run -it -ePORT=8080 -p8080:8080 sample-functions-framework-node

Python

docker run -it -ePORT=8080 -p8080:8080 sample-functions-framework-python

Ruby

docker run -it -ePORT=8080 -p8080:8080 sample-functions-framework-ruby

Visit the running function by browsing to localhost:8080.

Remote builds

Use Cloud Build to build your function into a container image, and Artifact Registry as the container repository to store and deploy each image.

Before you begin

Sign in to your Google Cloud account. If you're new to Google Cloud, create an account to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.

In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

Go to project selector

Make sure that billing is enabled for your Google Cloud project.

Enable the Cloud Build and Artifact Registry APIs.

Enable the APIs

Install the Google Cloud CLI.

To initialize the gcloud CLI, run the following command:

gcloud init

In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

Go to project selector

Make sure that billing is enabled for your Google Cloud project.

Enable the Cloud Build and Artifact Registry APIs.

Enable the APIs

Install the Google Cloud CLI.

To initialize the gcloud CLI, run the following command:

gcloud init

Ensure that your Google Cloud project has access to a container image repository.
To configure access to a Docker repository in Artifact Registry:
1. Create a new Docker repository in the same location of your Google Cloud project.
```
gcloud artifacts repositories create REPO_NAME \
--repository-format=docker \
--location=REGION --description="DESCRIPTION"
```
  Replace:
  - REPO_NAME with the name that you choose for your Docker repository.
  - REGION with the location in or nearest to the location of your Google Cloud project.
  - DESCRIPTION with a description of your choice.
  For example, to create a docker repository in us-west2 with the description "Docker repository", you run:
```
gcloud artifacts repositories create buildpacks-docker-repo --repository-format=docker \
--location=us-west2 --description="Docker repository"
```
2. Verify that your repository was created:
```
gcloud artifacts repositories list
```
  You should see name that you choose for your Docker repository in the list.

Build a function remotely

You use the gcloud builds submit command to build and upload your container image to your repository.

You can choose to specify your container image in the command itself or use a configuration file.

Build with command

To build without a configuration file, specify the image flag:

gcloud builds submit --pack image=LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/IMAGE_NAME

Replace:

LOCATION with the region name of your container repository, for example, us-west2
PROJECT_ID with the ID of your Google Cloud project.
REPO_NAME with the name of your Docker repository.
IMAGE_NAME with the name of your container image.

Example:

gcloud builds submit --pack image=us-west2-docker.pkg.dev/my-project-id/my-buildpacks-docker-repo

Build with configuration files

You can use a configuration file to define your image repository configuration details to simply the build command. The configuration file uses the YAML file format and must include a build step that uses the pack CLI.

Create a YAML file name cloudbuild.yaml that includes the URI of your container image repository.

  options:
    logging: CLOUD_LOGGING_ONLY
    pool: {}
  projectId: PROJECT_ID
  steps:
  - name: gcr.io/k8s-skaffold/pack
    entrypoint: pack
    args:
    - build
    - LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/IMAGE_NAME
    - --builder
    - gcr.io/buildpacks/builder:latest
    - --network
    - cloudbuild
  images:
  - LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/IMAGE_NAME

Replace:

LOCATION with the region name of your container repository, for example, us-west2.
PROJECT_ID with the ID of your Google Cloud project.
REPO_NAME with the name of your Docker repository.
IMAGE_NAME with the name of your container image.

Build the application.

If you named your configuration file cloudbuild.yaml, you can run the following command:
```
gcloud builds submit .
```
Tip: If you use a unique name for your configuration file, you can use the --config flag.

Example: Build a sample function remotely

The following examples demonstrate how to build a sample remotely, and verify that the container image was pushed to your repository in Artifact Registry.

Clone the sample repository to your local machine:

git clone https://github.com/GoogleCloudPlatform/buildpack-samples.git

Change to the directory that contains the application sample code:

Go

cd buildpack-samples/sample-functions-framework-go

Java

cd buildpack-samples/sample-functions-framework-java-mvn

Node.js

cd buildpack-samples/sample-functions-framework-node

Python

cd buildpack-samples/sample-functions-framework-python

Ruby

cd buildpack-samples/sample-functions-framework-ruby

Use gcloud to submit the application source code to Cloud Build:

Go

gcloud builds submit --pack image=LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/sample-functions-framework-go

Java

gcloud builds submit --pack image=LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/sample-functions-framework-gradle

Node.js

gcloud builds submit --pack image=LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/sample-functions-framework-node

Python

gcloud builds submit --pack image=LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/sample-functions-framework-python

Ruby

gcloud builds submit --pack image=LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME/sample-functions-framework-ruby

Replace:

LOCATION with the region name of your container repository. Example: us-west2-docker.pkg.dev
PROJECT_ID with the ID of your Google Cloud project.
REPO_NAME with the name of your Docker repository.

Verify that the sample function was successfully published into REPO_NAME:
```
gcloud artifacts docker images list LOCATION-docker.pkg.dev/PROJECT_ID/REPO_NAME
```
Replace:
- LOCATION with the region name of your container repository, for example, us-west2.
- PROJECT_ID with the ID of your Google Cloud project.
- REPO_NAME with the name of your Docker repository.

What's Next

Deploy your function image to Cloud Run.
Set environment variables.
Configure build images.
Speed up builds with cache images.