gcbuild

A super simple CLI and API for using Google Cloud Build.

Installation

npm install gcbuild

Command Line

gcb is a convenient way to submit jobs to Google Cloud Build. To use as a command line application:

npm install --save-dev gcbuild

Then from your package.json, it's super easy to add a deploy script:

"scripts": {
  "deploy": "gcb"
}

Positional arguments

SOURCE

Location of the sources to be deployed. If not specified, assumes the current working directory.

Flags

--config

The YAML or JSON file to use as the build configuration file. Defaults to 'cloudbuild.yaml' if not specified.

--tag

The tag to use with a "docker build" image creation.

--substitutions

Substitution variables to pass to the build. Can be specified multiple times. These are available in the build config as ${_VARIABLE_NAME}.

--timeout

Amount of time to wait for the build to complete. Examples: 20m, 1200s, 1h

--machine-type

Compute Engine machine type on which to run the build. Examples: E2_HIGHCPU_8, N1_HIGHCPU_32

Examples

# Create an image for the current working directory.
$ gcb

# If there's a Dockerfile in the CWD, I can also specify a tag
$ gcb --tag my-image-name

# Use a build file not named `cloudbuild.yaml`
$ gcb --config suchbuild.json

# Perform a build from another location on disk
$ gcb ~/Code/verydocker

# Use substitutions for template variables
$ gcb --substitutions=_ENV=prod --substitutions=_VERSION=1.2.3

# Set a custom timeout and machine type
$ gcb --timeout=20m --machine-type=E2_HIGHCPU_8

API

You can also use this as a regular old API.

Building

import {build} from 'gcbuild';

const result = await build({
  sourcePath: '/path/to/source',
  tag: 'my-image',
  substitutions: {
    _ENV: 'production',
    _VERSION: '1.2.3'
  },
  timeout: '20m',
  machineType: 'E2_HIGHCPU_8'
});
console.log('Build completed:', result.metadata.build.id);

Build Management

You can also list, retrieve, and cancel builds:

import {listBuilds, getBuild, cancelBuild} from 'gcbuild';

// List recent builds
const builds = await listBuilds({
  limit: 10,
  filter: 'status=WORKING'
});
console.log('Recent builds:', builds);

// Get a specific build
const build = await getBuild('build-id-123');
console.log('Build status:', build.status);

// Cancel a running build
const cancelled = await cancelBuild('build-id-123');
console.log('Cancelled:', cancelled.id);

Options

All methods accept the following authentication options:

• projectId: GCP project ID
• keyFilename: Path to service account key file
• Additional GoogleAuthOptions

Build-specific options for build():

• sourcePath: Path to the source code directory (default: CWD)
• configPath: Path to cloudbuild.yaml/json file
• tag: Docker image tag
• substitutions: Object with substitution variables for the build
• timeout: Build timeout (e.g., '20m', '1200s', '1h')
• machineType: Compute Engine machine type (e.g., 'E2_HIGHCPU_8')

Error Handling

The library throws BuildError exceptions with detailed information:

import {build, BuildError} from 'gcbuild';

try {
  await build({ sourcePath: './app' });
} catch (error) {
  if (error instanceof BuildError) {
    console.error('Build ID:', error.buildId);
    console.error('Status:', error.status);
    console.error('Failed step:', error.failedStep);
    console.error('Suggestion:', error.suggestion);
    console.error('Logs:', error.log);
  }
}

Authentication

This library uses google-auth-library under the hood to provide authentication. That means you can authenticate a few ways.

Using a service account

One of the reasons this library exists is to provide a nodejs native deployment in environments where you don't want to have the Cloud SDK installed.

For this method, you'll need to create a service account, and download a key.

1. In the GCP Console, go to the Create service account key page.
2. From the Service account drop-down list, select New service account.
3. In the Service account name field, enter a name.
4. From the Role drop-down list, select Project > Owner.
5. Click Create. A JSON file that contains your key downloads to your computer.

export GOOGLE_APPLICATION_CREDENTIALS="./keys.json"
gcb .

Using application default credentials

If you plan on only using this from your machine, and you have the Google Cloud SDK installed, you can just use application default credentials like this:

gcloud auth login
gcloud auth application-default login
gcloud config set project 'YOUR-AWESOME-PROJECT'
gcb .

License

MIT