Deploy Docker Image with Launchflow

Deploy a Docker Image to AWS Fargate with LaunchFlow.

View the source code for this guide in our examples repo.

0. Set up your Dockerfile

If you already have a Dockerfile you can skip to step #1.

Create a new directory for your project.

1mkdir launchflow-docker
2cd lauchflow-docker

Create a file named Dockerfile with the following content:

1. Initialize Launch Flow

Install the LaunchFlow Python SDK and CLI using pip.

1pip install launchflow[aws]

Initialize LaunchFlow in your project

1lf init --backend=local

Name your project
Select Yes for creating an example infra.py
Select AWS for your cloud provider
Select ECS Fargate for your service

Once finished you will get an infra.py that looks like:

1import launchflow as lf
2
3# ECSFargateService Docs: https://docs.launchflow.com/reference/aws-services/ecs-fargate
4api = lf.aws.ECSFargateService(
5    "my-ecs-api",
6    dockerfile="Dockerfile",  # Path to your Dockerfile
7)

ECSFargateService will build your Dockerfile and deploy to ECS Fargate. You can provide additional fields to ECSFargateService to configure things like machine type, num instances, or even a custom domain.

2. Deploy your Service

Make sure you have local AWS credentials set up before deploying.

1lf deploy

Name your environment (dev is a good first name)
Select your cloud provider AWS
Confirm the resources to be created
Select the service to deploy

Once complete you will see a link to your deployed service on ECS Fargate.

3. Cleanup your Resources

Optionally you can delete all your resources, service, and environments with:

1lf destroy
2lf environments delete

LaunchFlow Cloud usage is optional and free for individuals.

Using the local backend like we did above works fine for starting a project, but doesn't offer a way to share state between multiple users. LaunchFlow Cloud is a web-based service for managing, sharing, and automating your infrastructure. It's free small teams and provides a simple, secure way to collaborate with your team and automate your release pipelines.

1lf init --backend=lf

This will create a project in your LaunchFlow Cloud account and migrate your local state to the LaunchFlow Cloud backend.

What's next?

Add resources to your application
Promote your application to a production enviroment
Learn more about Environments, Services, and Resources
Join theLaunchFlow Slack community
View your application in theLaunchFlow console

0. Set up your Dockerfile

1. Initialize Launch Flow

2. Deploy your Service

3. Cleanup your Resources

4. Visualize, Share, and Automate

What's next?