Deploy to Fly.io

Fly.io is a platform for running applications close to users. This guide will show you how to deploy Flipt to Fly.io and configure Flipt to use a Postgres database, also managed by Fly.io.

​

What You’ll Learn

In this guide, you will learn how to deploy Flipt to Fly.io with Postgres. You’ll also learn how to configure Flipt with environment variables.

By the end of this guide, we will have:

• 🚀 Successfully deployed Flipt to Fly.io
• 🐘 Configured Flipt with environment variables to use Fly.io’s managed Postgres service

​

Prerequisites

• A Fly.io account (Sign up: https://fly.io)
• flyctl CLI installed on your local machine (Installation guide: https://fly.io/docs/getting-started/installing-flyctl/)

​

Deployment Steps

1. Ensure you can log in to your Fly.io account with fly auth login.
2. Create and cd into a new directory for your Flipt deployment. (e.g. mkdir flipt-test && cd flipt-test)
3. Begin the launch process to deploy Flipt on Fly.io using their CLI, selecting mainly the defaults and Postgres when prompted:

$ fly launch -i ghcr.io/flipt-io/flipt:latest

? App Name (leave blank to use an auto-generated name) flipt-test

Automatically selected personal organization: <your-org>

...

? Would you like to set up a Postgresql database now? Yes

4. After the launch process completes, it will write a fly.toml file to your current directory. You can configure the number of instances, memory, and CPU allocated to your Flipt deployment. You can also set environment variables to customize Flipt’s configuration.

   For more information, refer to the Fly.io documentation and the Flipt documentation.

5. Before deploying Flipt, we’ll need to set the FLIPT_DB_URL secret to point to your newly configured Postgres database.

   You can do this with the following command, replacing <your-db-url> with the connection string from the launch process and appending the db name (postgres) and ?sslmode=disable:

fly secrets set FLIPT_DB_URL=<your-db-url>/postgres?sslmode=disable

6. Finally, deploy Flipt to Fly.io with fly deploy.

$ fly deploy

Deploying flipt-test
==> Validating App Configuration

​

Verifying the Deployment

After deployment, you can verify if Flipt is running correctly by accessing the Flipt UI on the provided Fly.io URL or issue the command fly open in the CLI to open your newly deployed Flipt instance in the browser.

$ fly open

opening https://flipt-test.fly.dev/ ...

​

Configuration

You can configure Flipt with environment variables in the fly.toml file. For example, to configure Flipt to use a custom port and enable DEBUG logging you can add the following to the fly.toml file:

# See https://fly.io/docs/reference/configuration/ for information
# about how to use this file.

app = "flipt-test"
primary_region = "iad"

[build]
  image = "ghcr.io/flipt-io/flipt:latest"

[http_service]
  internal_port = 9090
  force_https = true

[env]
  FLIPT_SERVER_HTTP_PORT = 9090
  FLIPT_LOG_LEVEL = "debug"

After making changes to the fly.toml file, you can deploy the changes with fly deploy.

$ fly deploy

Deploying flipt-test
==> Validating App Configuration

​

Troubleshooting

If you encounter any issues during the deployment, check the logs on Fly.io for any error messages:

fly logs

Enabling DEBUG logging as shown above can be helpful for troubleshooting any issues during Flipt startup.

​

Conclusion

Deploying Flipt to Fly.io allows you to get up and running with Flipt quickly. For production deployments however, you’ll likely want to configure Flipt with authentication as well as consider configuring caching, observability, and using a read replica for your database.

For more information on production deployments, refer to the Deployment section of the documentation.