Starting builds manually

This page describes how to start a build in Cloud Build manually using the gcloud command-line tool and the Cloud Build API.

Before you begin

  • If you want to use the command-line examples in this guide, install the gcloud command-line tool.
  • To build using the Cloud Build build config, build using a build config file
  • To build using a Dockerfile, have your Dockerfile ready.
  • If your build requires source code, have your source code ready.

Required IAM permissions

For instructions on granting IAM roles, see Configure access for project members.

Running builds

Console

You cannot manually start a new build using Google Cloud Console, however you can retry a previous build.

To retry a previous build:

  1. Open the Build History page in the Cloud Build section in the Google Cloud Console.

    Open the Build History page

  2. Select your project and click Open.

  3. In the Build history page, click on a build that you wish to rebuild.

  4. Click Rebuild.

gcloud

Using a Dockerfile:

Your Dockerfile contains all information needed to build a Docker image using Cloud Build.

To run a build request using your Dockerfile, run the following command from the directory containing your application code, Dockerfile, and any other assets:

gcloud builds submit --tag gcr.io/project-id/image-name .

Where:

  • project-id is the name of your Cloud project.
  • image-name is the image to be built.
  • . specifies that the source code is in the current working directory.

The full name of the image to be built is `gcr.io/project-id/image-name. Images pushed to Container Registry use the registry name format.

The gcloud builds submit command:

  • compresses your application code, Dockerfile, and any other assets in the current directory as indicated by .;
  • uploads the files to a Cloud Storage bucket;
  • initiates a build using the uploaded files as input;
  • tags the image using the provided name
  • pushes the built image to Container Registry.

As the build progresses, its output is displayed in your shell or terminal window. When the build is complete, you should see an output similar to the following:

    DONE
    ---------------------------------------------------------------------------------
    ID                                    CREATE_TIME                DURATION STATUS
    $BUILD_ID                             2016-10-28T15:21:18+00:00  12S      SUCCESS

where $BUILD_ID is your build's unique identifier.

Using the Cloud Build build config file:

To submit a build using the build config, run the following command:

    gcloud builds submit --config build-config source-code

where:

  • build-config is the path to the build config file.
  • source-code is the path or URL source code.

For example, the following command submits the cloudbuild.yaml build request using archived source code stored in a Cloud Storage bucket.

    gcloud builds submit --config cloudbuild.yaml \
        gs://cloud-build-examples/node-docker-example.tar.gz

You can use . to specify that the source code is in the current working directory:

    gcloud builds submit --config cloudbuild.yaml .

gcloudignore: When including source code for the build, the above command uploads all of the files in the specified directory to Google Cloud Platform to build. If you want to exclude certain files in the directory, you can include a file named .gcloudignore in the top-level upload directory; the files that it specifies will be ignored. If no .gcloudignore file is present in the top-level upload directory, but a .gitignore file is, the gcloud tool will generate a Git-compatible .gcloudignore file that respects your .gitignore-ed files. For more information, see the gcloudignore documentation.

If you do not have source code to pass in to your build, use the --no-source flag where build-config is the path to the build config file:

    gcloud builds submit --config build-config --no-source

API

To submit the build request using curl:

  1. Create a file named request.json with the following contents:

    {
        "source": {
            "storageSource": {
                "bucket": "cloud-build-examples",
                "object": "node-docker-example.tar.gz"
            }
        },
        "steps": [{
            "name": "gcr.io/cloud-builders/docker",
            "args": [
                "build",
                "-t",
                "gcr.io/$PROJECT_ID/my-image",
                "."
            ]
        }],
        "images": [
            "gcr.io/$PROJECT_ID/my-image"
        ]
    }
    

    In this build request, Cloud Build calls the docker build step with the arguments build -t gcr.io/$PROJECT_ID/cb-demo-img ..

    The full name of the image to be built is gcr.io/$PROJECT_ID/cb-demo-img. Images pushed to Container Registry use the registry name format.

    The code source for the build is in a compressed tar archive, node-docker-example.tar.gz. The file is stored in a Cloud Storage bucket named cloud-build-examples.

  2. Run the following command where project-id is your Cloud project ID:

    curl -X POST -T request.json -H "Authorization: Bearer $(gcloud config config-helper \
        --format='value(credential.access_token)')" \
        https://cloudbuild.googleapis.com/v1/projects/project-id/builds
    

    In this command, curl sends request.json in a POST call to the builds endpoint for the projects.builds.create API method.

    The command displays details about your build in your shell or terminal window. The output is a JSON response and appears similar to the following:

    {
        "name": "operations/build/$PROJECT-ID/NmZhZW...",
        "metadata": {
            "@type": "type.googleapis.com/google.devtools.cloudbuild.v1.BuildOperationMetadata",
            "build": {
                "id": $BUILD-ID,
                "status": "QUEUED",
                "source": {
                    "storageSource": {
                        "bucket": "cloud-build-examples",
                        "object": "node-docker-example.tar.gz"
                    }
                },
                "createTime": "2017-05-12T18:58:07.341526Z",
                "steps": [
                {
                    "name": "gcr.io/cloud-builders/docker",
                    "args": [
                        "build",
                        "-t",
                        "gcr.io/$PROJECT-ID/cb-demo-img",
                        "."
                    ]
                }
                ],
                "timeout": "600s",
                "images": [
                    "gcr.io/$PROJECT-ID/cb-demo-img"
                ],
                "projectId": $PROJECT-ID,
                "logsBucket": "gs://...",
                "sourceProvenance": {
                    "resolvedStorageSource": {
                        "bucket": "cloud-build-examples",
                        "object": "node-docker-example.tar.gz",
                        "generation": "..."
                    }
                },
                "logUrl": "https://console.cloud.google.com/gcr/builds/...?project=$PROJECT-ID"
            }
        }
    }
    

    The JSON response is modeled using the Operation resource in the Cloud Build API. The metadata field is modeled using the Build resource. The QUEUED status indicates that the build is awaiting execution.

What's next