Infrastructure as code made fast & easy

The core of Terraforge lives in @terraforge/core, with the Terraform bridge in @terraforge/terraform. Together they provide a fast, diff-friendly IaC workflow for AWS (and custom) stacks.

The problem

The most used IaC solutions are slow & don't effectively leverage diffing to speed up their deployments.

Setup

Install with (NPM):

npm i @terraforge/core @terraforge/terraform @terraforge/aws

Example

First, you need to create a workspace instance and pass in the cloud providers that you will use. We also need to give it a lock backend & state backend.

• A cloud provider is used to create resources on a specific cloud provider. We have built-in cloud providers for AWS resources, but you could simply add your own as well.

• The state backend is used for storing the latest deployment state.

• The lock backend is used for acquiring a lock when you deploy your app. This will guarantee that multiple people can never deploy the same application at the same time.

• The activity log backend is used for storing the activity log when every you deploy / delete your app.

In this example, we will use a local file lock & state provider.

import { App, FileStateBackend, FileLockBackend, Stack, WorkSpace } from '@terraforge/core'
import { aws } from '@terraforge/aws'

const workspace = new WorkSpace({
	providers: [
		aws({
			profile: "PROFILE",
			region: "REGION"
		})
	],
	backend: {
		state: new FileStateBackend({ dir: './states' }),
		lock: new FileLockBackend({ dir: './locks' }),
	}
})

With your workspace configuration ready we can now move on to defining your infrastructure. This example illustrates how simple it is to define multi-stack resources without worrying about cross-stack references.

const app = new App('todo-app')
const storage = new Stack(app, 'storage')
const list = new aws.s3.Bucket(storage, 'todo-list', {
	bucket: 'your-bucket-name'
})

const items = new Stack(app, 'items')
const todo = new aws.s3.BucketObject(items, 'todo-item', {
	bucket: list.bucket,
	key: 'item-1',
	content: JSON.stringify({
		title: 'Write docs...',
		done: true,
	}),
})

After defining your infrastructure, we can deploy our app.

await workspace.deploy(app)

Or destroy our app.

await workspace.delete(app)

Maybe you want to only deploy a subset of stacks.

await workspace.deploy(app, {
	filters: ["storage"]
})

Production

For production, we recommend you use a state & lock backend that stores your data in the cloud. An AWS DynamoDB table is perfect for storing locks. While AWS S3 is perfect for storing your state files.

const workspace = new WorkSpace({
	providers: [...],
	backend: {
		lock: new DynamoLockBackend({
  			region,
  			credentials,
  			tableName: 'awsless-locks',
  		}),
		state: new S3StateBackend({
			region,
			credentials,
			bucket: 'awsless-state-UNIQUE_ID',
		}),
	}
})