Static Site Automation
What it is?
This Gihub action that uses the golang aws sdk to build s3 bucket and attach policy
for static website deploy the static file to that newly created s3 bucket and
comment the url to the PR. To deploy static file it uses either from your
repository or build during your workflow. There is self hosted tool, if Github
action is not feasible.
Best for?
- Immediate feedback visually to developers or anyone interested in changes.
- Reduce burden of having to build application for QA and verify the changes.
- Faster iterations.
How to use?
Add .yml
file/s such as given examples in your .github/workflows
folder. Refer to the documentation on workflow YAML syntax here.
The following example will:
- Create s3 bucket and attach policy for static site
- Build the javascript/typescript frontend application with the help of
given command (ex:
BUILD_COMMAND="yarn build"
)
- Upload build file (static site) to s3
- Comment the URL of the static site to the Pull Request
- Delete the aws S3 bucket after PR is merged
Config file: .github/workflows/deploy-existing.yml
name: Next js frontend dev
on:
push:
branches:
- dev
jobs:
deploy:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@master
- name: Build and deploy Studio app
uses: razzkumar/pr-automation-with-s3@v1.0.2
env:
AWS_S3_BUCKET: ${{ secrets.AWS_S3_BUCKET }}
AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
AWS_REGION: "us-east-1"
SRC_FOLDER: "out"
ACTION: 'deploy'
BUILD_COMMAND: "yarn build && yarn export"
CLOUDFRONT_ID: ${{ secrets.CLOUDFRONT_ID }}
SECRETS_MANAGER: ${{ secrets.SECRETS_MANAGER }} // name of secrets on secret manager
Config file: .github/workflows/deploy-on-pr.yml
name: Deploy site to S3 And add comment to PR and delete after merge
on:
pull_request:
branches:
- master
jobs:
deploy:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@master
- name: Static site deploy to s3 and comment on PR
uses: razzkumar/pr-automation-with-s3@v1.0.2
env:
AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
GH_ACCESS_TOKEN: ${{ secrets.GH_ACCESS_TOKEN}}
AWS_REGION: 'us-east-2' # optional: defaults to us-east-2
SRC_FOLDER: 'build' # optional: defaults to build (react app)
IS_BUILD: 'true' # optional: defaults to true
ACTION: "create" # optional: defaults to create (option:create,delete and deploy)
BUILD_COMMAND: "yarn build" # optional: defaults to `yarn build`
Config file: .github/workflows/cleanup-on-pr-merge.yml
name: Delete S3 bucket after PR merge
on:
pull_request:
types: [closed]
jobs:
delete:
runs-on: ubuntu-latest
steps:
- name: Clean up temperory bucket
if: github.event.pull_request.merged == true
uses: razzkumar/pr-automation-with-s3@v1.0.2
env:
AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
AWS_REGION: 'us-east-2' # optional: defaults to us-east-2
ACTION: "delete" # Action must be delete to delete
- It can be used for many other purpose also,some of the examples are:
Configuration
The following settings must be passed as environment variables as shown in the example. Sensitive information, especially GH_ACCESS_TOKEN
,AWS_ACCESS_KEY_ID
and AWS_SECRET_ACCESS_KEY
, should be set as encrypted secrets — otherwise, they'll be public to anyone browsing your repository's source code and CI logs.
Key |
Suggested Type |
Value |
Required |
Default |
GH_ACCESS_TOKEN |
secrect env |
Your Github access token used while commenting PR |
YES/NO If ACTION: create then it's required,otherwise it's optional |
NA |
AWS_ACCESS_KEY_ID |
secret env |
Your AWS Access Key. More info here. |
Yes |
N/A |
AWS_SECRET_ACCESS_KEY |
secret env |
Your AWS Secret Access Key. More info here. |
Yes |
N/A |
AWS_S3_BUCKET |
secret env |
The name of the bucket you're syncing to. For example, jarv.is or my-app-releases . |
YES/NO |
- If running on PR it will genereat by tool PR-Branch .prPR-number .auto-deploy - In the case of depoyment it required |
AWS_REGION |
env |
The region where you created your bucket. Set to us-east-2 by default. Full list of regions here. |
No |
us-east-2 |
SRC_FOLDER |
env |
The local directory (or file) you wish to deploy to S3. For example, public . Defaults to build . |
No |
build (based on react app) |
IS_BUILD |
env |
This is the flag that indicate that build a project or not |
No |
true (It will run yarn && yarn build by default) |
ACTION |
env |
This is also a flag that indicate what to do (create :-create s3 (if not exist) bucket,build react and comment on PR,deploy :helps to deploy to s3,delete : delete the s3 bucket) |
No |
create (It will create s3 (if not exist),built the app, deploy to s3 and comment URL to PR`) |
BUILD_COMMAND |
env |
How to build the react app if its npm run build then it will run npm install && npm run build |
No |
yarn build (It will run yarn && yarn build by default) |
CLOUDFRONT_ID |
secret env |
id of cloudfront for invalidation |
No |
|
SECRETS_MANAGER |
env |
name of the aws secres manager key |
No |
|
Note for S3 Bucket creation
- It only create a s3 bucket if not
exist
- While Creating bucket for the pull_request S3 bucket name will be:
PR-Branch
.prPR-number
.auto-deploy
- For Eg.:
- if base branch is
SIG-1000
and PR number is 23
the the bucket name will be sig-100.pr23.auto-deploy
- If we deploy app on push or (not on pull requst) like prebuild app deployment, app build and deploy then the bucket name will be
$AWS_S3_BUCKET.auto-deploy
- For Eg.
- if
AWS_S3_BUCKET=dev-test-deployment
then bucket will be dev-test-deployment.auto-deploy
TODO
- Add tests
- Add option to deploy on aws cloudfront
- Design PR comment done by tool
- Maintain code quality
Contributing
Feel free to send pull requests
License
This project is distributed under the MIT license