deploy-cloudrun

The deploy-cloudrun GitHub Action deploys to Google Cloud Run. It can deploy a container image or from source, and the resulting service URL is available as a GitHub Actions output for use in future steps.

This is not an officially supported Google product, and it is not covered by a Google Cloud support contract. To report bugs or request features in a Google Cloud product, please contact Google Cloud support.

Prerequisites

• This action requires Google Cloud credentials that are authorized to access the secrets being requested. See Authorization for more information.

• This action runs using Node 20. If you are using self-hosted GitHub Actions runners, you must use a runner version that supports this version or newer.

Usage

jobs:
  job_id:
    # ...

    permissions:
      contents: 'read'
      id-token: 'write'

    steps:
    - uses: 'actions/checkout@v4'

    - uses: 'google-github-actions/auth@v2'
      with:
        workload_identity_provider: 'projects/123456789/locations/global/workloadIdentityPools/my-pool/providers/my-provider'
        service_account: '[email protected]'

    - id: 'deploy'
      uses: 'google-github-actions/deploy-cloudrun@v2'
      with:
        service: 'hello-cloud-run'
        image: 'gcr.io/cloudrun/hello'

    - name: 'Use output'
      run: 'curl "${{ steps.deploy.outputs.url }}"'

Inputs

• service: (Required, unless providing metadata or job) ID of the service or fully-qualified identifier of the service. Only one of service or job may be specified.

• job: (Required, unless providing metadata or service) ID of the job or fully-qualified identifier of the job. Only one of service or job may be specified.

• image: (Required, unless providing metadata or source) Fully-qualified name of the container image to deploy. For example:

  gcr.io/cloudrun/hello:latest
  

  or

  us-docker.pkg.dev/my-project/my-container/image:1.2.3
  

• source: (Required, unless providing metadata, image, or job) Path to source to deploy. If specified, this will deploy the Cloud Run service from the code specified at the given source directory.

  This requires the Artifact Registry API to be enabled. Furthermore, the deploying service account must have the Cloud Build Service Account role. The initial deployment will create an Artifact Registry repository which requires the Artifact Registry Admin role.

  Learn more about Deploying from source code.

• suffix: (Optional) String suffix to append to the revision name. Revision names always start with the service name automatically. For example, specifying 'v1' for a service named 'helloworld', would lead to a revision named 'helloworld-v1'. The default value is no suffix.

• env_vars: (Optional) List of key=value pairs to set as environment variables. All existing environment variables will be retained. If both env_vars and env_vars_file are specified, the keys in env_vars will take precendence over the keys in env_vars_files.

  with:
    env_vars: |
      FOO=bar
      ZIP=zap

  Entries are separated by commas (,) and newline characters. Keys and values are separated by =. To use ,, =, or newline characters, escape them with a backslash:

  with:
    env_vars: |
      [email protected]\,[email protected]

• env_vars_file: (Optional) Path to a file on disk, relative to the workspace, that defines environment variables. The file can be newline-separated KEY=VALUE pairs, JSON, or YAML format. If both env_vars and env_vars_file are specified, the keys in env_vars will take precendence over the keys in env_vars_files.

  FOO=bar
  ZIP=zap
  

  or

  {
    "FOO": "bar",
    "ZIP": "zap"
  }

  or

  FOO: 'bar'
  ZIP: 'zap'

  When specified as KEY=VALUE pairs, the same escaping rules apply as described in env_vars. You do not have to escape YAML or JSON.

• secrets: (Optional) List of key=value pairs to use as secrets. These can either be injected as environment variables or mounted as volumes. All existing environment secrets and volume mounts will be retained.

  with:
    secrets: |-
      # As an environment variable:
      KEY1=secret-key-1:latest
  
      # As a volume mount:
      /secrets/api/key=secret-key-2:latest

  The same rules apply for escaping entries as from env_vars, but Cloud Run is more restrictive with allowed keys and names for secrets.

• labels: (Optional) List of key=value pairs to set as labels on the Cloud Run service. Existing labels will be overwritten.

  with:
    labels: |-
      my-label=my-value

  The same rules apply for escaping entries as from env_vars, but labels have strict naming and casing requirements. See Requirements for labels for more information.

• skip_default_labels: (Optional) Skip applying the special annotation labels that indicate the deployment came from GitHub Actions. The GitHub Action will automatically apply the following labels which Cloud Run uses to enhance the user experience:

  managed-by: github-actions
  commit-sha: <sha>
  

  Setting this to true will skip adding these special labels. The default value is false.

• tag: (Optional) Traffic tag to assign to the newly-created revision.

• timeout: (Optional) Maximum request execution time, specified as a duration like "10m5s" for ten minutes and 5 seconds.

• flags: (Optional) Space separate list of other Cloud Run flags. This can be used to access features that are not exposed via this GitHub Action.

  with:
    flags: '--add-cloudsql-instances=...'

  See the complete list of flags for more information.

  Please note, this GitHub Action does not parse or validate the flags. You are responsible for making sure the flags are available on the gcloud version and subcommand. When using tag_traffic or revision_traffic, the subcommand is gcloud run services update-traffic. For all other values, the subcommand is gcloud run deploy.

• no_traffic: (Optional) If true, the newly deployed revision will not receive traffic. The default value is false.

• revision_traffic: (Optional, mutually-exclusive with tag_traffic) Comma-separated list of revision traffic assignments.

  with:
    revision_traffic: 'my-revision=10' # percentage

  To update traffic to the latest revision, use the special tag "LATEST":

  with:
    revision_traffic: 'LATEST=100'

• tag_traffic: (Optional, mutually-exclusive with revision_traffic) Comma-separated list of tag traffic assignments.

  with:
    tag_traffic: 'my-tag=10' # percentage

• project_id: (Optional) ID of the Google Cloud project in which to deploy the service. The default value is computed from the environment.

• region: (Optional) Regions in which the Cloud Run services are deployed. This can be a single region or a comma-separated list of regions. The default value is us-central1.

• gcloud_version: (Optional) Version of the gcloud CLI to use. The default value is latest.

• gcloud_component: (Optional) Component of the gcloud CLI to use. Valid values are alpha and beta.

Custom metadata YAML

For advanced use cases, you can define a custom Cloud Run metadata file. This is a YAML description of the Cloud Run service. This allows you to customize your service configuration, such as memory limits, CPU allocation, max instances, and more.

⚠️ When using a custom metadata YAML file, all other inputs are ignored!

• metadata: (Optional) The path to a Cloud Run service metadata file.

To deploying a new service to create a new YAML service definition:

apiVersion: serving.knative.dev/v1
kind: Service
metadata:
  name: SERVICE
spec:
  template:
    spec:
      containers:
      - image: IMAGE

To update a revision or to deploy a new revision of an existing service, download and modify the YAML service definition:

gcloud run services describe SERVICE --format yaml > service.yaml

Allowing unauthenticated requests

A Cloud Run product recommendation is that CI/CD systems not set or change settings for allowing unauthenticated invocations. New deployments are automatically private services, while deploying a revision of a public (unauthenticated) service will preserve the IAM setting of public (unauthenticated). For more information, see Controlling access on an individual service.

Outputs

• url: The URL of your Cloud Run service.

Authorization

There are a few ways to authenticate this action. The caller must have permissions to access the secrets being requested.

You will need to authenticate to Google Cloud as a service account with the following roles:

• Cloud Run Admin (roles/run.admin):
  • Can create, update, and delete services.
  • Can get and set IAM policies.

This service account needs to be a member of the Compute Engine default service account, ([email protected]), with role Service Account User. To grant a user permissions for a service account, use one of the methods found in Configuring Ownership and access to a service account.

Via google-github-actions/auth

Use google-github-actions/auth to authenticate the action. You can use Workload Identity Federation or traditional Service Account Key JSON authentication.

jobs:
  job_id:
    permissions:
      contents: 'read'
      id-token: 'write'

    steps:

    # ...

    - uses: 'google-github-actions/auth@v2'
      with:
        workload_identity_provider: 'projects/123456789/locations/global/workloadIdentityPools/my-pool/providers/my-provider'
        service_account: '[email protected]'

    - uses: 'google-github-actions/deploy-cloudrun@v2'
      with:
        image: 'gcr.io/cloudrun/hello'
        service: 'hello-cloud-run'

Via Application Default Credentials

If you are hosting your own runners, and those runners are on Google Cloud, you can leverage the Application Default Credentials of the instance. This will authenticate requests as the service account attached to the instance. This only works using a custom runner hosted on GCP.

jobs:
  job_id:
    steps:
    # ...

    - uses: 'google-github-actions/deploy-cloudrun@v2'
      with:
        image: 'gcr.io/cloudrun/hello'
        service: 'hello-cloud-run'

The action will automatically detect and use the Application Default Credentials.

Example Workflows

• Deploy from source

• Build and deploy a container