A GitHub Action that starts live environments for your pull requests and branches.
Once installed in your repository, this action is triggered any time a change
is made to Pull Requests labelled with the pullpreview
label, or one of the
always-on branches.
When triggered, it will:
- Check out the repository code
- Provision a cheap AWS Lightsail instance, with docker and docker-compose set up
- Continuously deploy the specified pull requests and branches, using your docker-compose file(s)
- Report the preview instance URL in the GitHub UI
It is designed to be the no-nonsense, cheap, and secure alternative to services that require access to your code and force your app to fit within their specific deployment system and/or require a specific config file.
- Product Owners: Interact with a new feature as it's built, give valuable feedback earlier, reduce wasted development time.
- Developers: Show your work in progress, find bugs early, deliver the right feature.
- Ops: Concentrate on high value tasks, not maintaining staging environments.
- CTOs: Don't let your code run on third-party servers: your code always stays private on either GitHub's or your servers.
Preview environments that:
-
work with your existing tooling: If your app can be started with docker-compose, it can be deployed to preview environments with PullPreview.
-
can be started and destroyed easily: You can manage preview environments by adding or removing the
pullpreview
label on your Pull Requests. You can set specific branches as always on, for instance to continuously deploy your master branch. -
are cheap too run: Preview environments are launched on AWS Lightsail instances, which are both very cheap (10USD per month, proratized to the duration of the PR), and all costs included (bandwith, storage, etc.)
-
take the privacy of your code seriously: The workflow happens all within a GitHub Action, which means your code never leaves GitHub or your Lightsail instances.
-
make the preview URL easy to find for your reviewers: Deployment statuses and URLs are visible in the PR checks section, and on the Deployments tab in the GitHub UI.
-
persist state across deploys: Any state stored in docker volumes (e.g. database data) will be persisted across deploys, making the life of reviewers easier.
-
are easy to troubleshoot: You can give specific GitHub users the authorization to SSH into the preview instance (with sudo privileges) to further troubleshoot any issue. The SSH keys that they use to push to GitHub will automatically be installed on the preview servers.
-
are integrated into the GitHub UI: Logs for each deployment run are available within the Actions section, and direct links to the preview environments are available in the Checks section of a PR, and in the Deployments tab of the repository.
- Getting Started
- Action Inputs / Outputs
- Handling Seed Data
- Workflow Examples
- FAQ
→ Please see the wiki for the full documentation.
Workflow file with the master
branch always on:
# .github/workflows/pullpreview.yml
name: PullPreview
on:
# the schedule is optional, but helps to make sure no dangling resources are left when GitHub Action does not behave properly
schedule:
- cron: "30 2 * * *"
push:
branches:
- main
pull_request:
types: [labeled, unlabeled, synchronize, closed, reopened]
jobs:
deploy:
if: github.event_name == 'schedule' || github.event_name == 'push' || github.event.label.name == 'pullpreview' || contains(github.event.pull_request.labels.*.name, 'pullpreview')
runs-on: ubuntu-latest
timeout-minutes: 30
steps:
- uses: actions/checkout@v4
- uses: pullpreview/action@v5
with:
# Those GitHub users will have SSH access to the servers
admins: crohr,other-github-user
# A preview environment will always exist for the main branch
always_on: main
# Use the cidrs option to restrict access to the live environments to specific IP ranges
cidrs: "0.0.0.0/0"
# PullPreview will use those 2 files when running docker-compose up
compose_files: docker-compose.yml,docker-compose.pullpreview.yml
# The preview URL will target this port
default_port: 80
# Use a 512MB RAM instance type instead of the default 2GB
instance_type: nano
# Ports to open on the server
ports: 80,5432
env:
AWS_ACCESS_KEY_ID: "${{ secrets.AWS_ACCESS_KEY_ID }}"
AWS_SECRET_ACCESS_KEY: "${{ secrets.AWS_SECRET_ACCESS_KEY }}"
AWS_REGION: "us-east-1"
The code for this Action is completely open source, and licensed under the Prosperity Public License (see LICENSE).
If you are a non-profit individual, then it is free to run (in that case, please tell me so and/or pass the word around!).
In all other cases, you must buy a license. More details on pullpreview.com.
Thanks for reading until the end!