GitHub Action
PR Review Apps on fly.io
This GitHub action wraps the Fly.io CLI to automatically deploy pull requests to fly.io for review. These are useful for testing changes on a branch without having to setup explicit staging environments.
This action will create, deploy, and destroy Fly apps. Just set an Action Secret for FLY_API_TOKEN.
If you have an existing fly.toml in your repo, this action will copy it with a new name when deploying. By default, Fly apps will be named with the scheme pr-{number}-{repo_org}-{repo_name}.
| name | description |
|---|---|
name |
The name of the Fly app. Alternatively, set the env FLY_APP. For safety, must include the PR number. Example: myapp-pr-${{ github.event.number }}. Defaults to pr-{number}-{repo_org}-{repo_name}. |
image |
Optional pre-existing Docker image to use |
config |
Optional path to a custom Fly toml config. Config path should be relative to path parameter, if specified. |
region |
Which Fly region to run the app in. Alternatively, set the env FLY_REGION. Defaults to iad. |
org |
Which Fly organization to launch the app under. Alternatively, set the env FLY_ORG. Defaults to personal. |
path |
Path to run the flyctl commands from. Useful if you have an existing fly.toml in a subdirectory. |
postgres |
Optional name of an existing Postgres cluster to flyctl postgres attach to. |
update |
Whether or not to update this Fly app when the PR is updated. Default true. |
secrets |
Secrets to be set on the app. Separate multiple secrets with a space |
vmsize |
Set app VM to a named size, eg. shared-cpu-1x, dedicated-cpu-1x, dedicated-cpu-2x etc. Takes precedence over cpu, cpu kind, and memory inputs. |
cpu |
Set app VM CPU (defaults to 1 cpu). Default 1. |
cpukind |
Set app VM CPU kind - shared or performance. Default shared. |
memory |
Set app VM memory in megabytes. Default 256. |
ha |
Create spare machines that increases app availability. Default false. |
FLY_API_TOKEN - Required. The token to use for authentication. You can find a token by running flyctl auth token or going to your user settings on fly.io.
name: Staging App
on:
pull_request:
types: [opened, reopened, synchronize, closed]
env:
FLY_API_TOKEN: ${{ secrets.FLY_API_TOKEN }}
FLY_REGION: iad
FLY_ORG: personal
jobs:
staging_app:
runs-on: ubuntu-latest
# Only run one deployment at a time per PR.
concurrency:
group: pr-${{ github.event.number }}
# Create a GitHub deployment environment per staging app so it shows up
# in the pull request UI.
environment:
name: pr-${{ github.event.number }}
url: ${{ steps.deploy.outputs.url }}
steps:
- uses: actions/checkout@v4
- name: Deploy
id: deploy
uses: superfly/[email protected]This action will destroy the Fly app, but it will not destroy the GitHub environment, so those will hang around in the GitHub UI. If this is bothersome, use an action like strumwolf/delete-deployment-environment to delete the environment when the PR is closed.
on:
pull_request:
types: [opened, reopened, synchronize, closed]
# ...
jobs:
staging_app:
# ...
# Create a GitHub deployment environment per review app.
environment:
name: pr-${{ github.event.number }}
url: ${{ steps.deploy.outputs.url }}
steps:
- uses: actions/checkout@v4
- name: Deploy app
id: deploy
uses: superfly/[email protected]
- name: Clean up GitHub environment
uses: strumwolf/delete-deployment-environment@v2
if: ${{ github.event.action == 'closed' }}
with:
# ⚠️ The provided token needs permission for admin write:org
token: ${{ secrets.GITHUB_TOKEN }}
environment: pr-${{ github.event.number }}If you have an existing Fly Postgres cluster you can attach it using the postgres action input. flyctl postgres attach will be used, which automatically creates a new database in the cluster named after the Fly app and sets DATABASE_URL.
For production apps, it's a good idea to create a new Postgres cluster specifically for staging apps.
# ...
steps:
- uses: actions/checkout@v4
- name: Deploy app
id: deploy
uses: superfly/[email protected]
with:
postgres: myapp-postgres-staging-appsIf you need to run multiple Fly apps per staging app, for example Redis, memcached, etc, just give each app a unique name. Your application code will need to be able to discover the app hostnames.
Redis example:
steps:
- uses: actions/checkout@v4
- name: Deploy redis
uses: superfly/[email protected]
with:
update: false # Don't need to re-deploy redis when the PR is updated
path: redis # Keep fly.toml in a subdirectory to avoid confusing flyctl
image: flyio/redis:6.2.6
name: pr-${{ github.event.number }}-myapp-redis
- name: Deploy app
id: deploy
uses: superfly/[email protected]
with:
name: pr-${{ github.event.number }}-myapp-app