Lab: deploy from CI

You need A GitHub account, csph signed in, and your lab project and environment IDs (lesson 5.1.1).

1. Fork the sample and turn on Actions

   Fork computesphere-nodejs-example, a small Express app with a Dockerfile, on port 3000. GitHub turns Actions off in forks: open the fork's Actions tab and enable workflows.

   You should seeYour fork's Actions tab lets you run workflows.

2. Create a ComputeSphere token

   In the console: Settings, Accounts, your account, Tokens, New token. Name it gh-actions-deploy, choose Project access and your lab project, set an expiry, choose Create token. It's shown once.

   You should seeA token starting csph_, copied.

3. Create a GitHub pull token

   On GitHub: your profile's Settings, Developer settings, Personal access tokens, Tokens (classic). Generate one with only read:packages.

   You should seeA classic token with only read:packages, copied.

4. Store both in the fork

   In the fork: Settings, Secrets and variables, Actions. Add secrets COMPUTESPHERE_API_TOKEN and GHCR_PULL_TOKEN, and variables COMPUTESPHERE_PROJECT_ID and COMPUTESPHERE_ENV_ID.

   You should seeTwo secrets and two variables listed.

5. Add the workflow and commit to main

   In the fork on GitHub, choose Add file, then Create new file, name it .github/workflows/deploy.yaml, paste this, and commit to main:

   Copy

   name: Deploy
   on:
     push:
       branches: [main]
     workflow_dispatch:
   permissions:
     contents: read
     packages: write
   jobs:
     deploy:
       runs-on: ubuntu-latest
       steps:
         - uses: actions/checkout@v4
         - id: image
           run: echo "ref=ghcr.io/${GITHUB_REPOSITORY,,}:${GITHUB_SHA::12}" >> "$GITHUB_OUTPUT"
         - uses: docker/login-action@v3
           with:
             registry: ghcr.io
             username: ${{ github.actor }}
             password: ${{ secrets.GITHUB_TOKEN }}
         - uses: docker/build-push-action@v6
           with:
             context: .
             push: true
             tags: ${{ steps.image.outputs.ref }}
         - id: deploy
           uses: computesphere/deploy@v1
           with:
             token: ${{ secrets.COMPUTESPHERE_API_TOKEN }}
             image: ${{ steps.image.outputs.ref }}
             name: nodejs-ci
             port: "3000"
             registry-url: https://ghcr.io
             registry-username: ${{ github.actor }}
             registry-password: ${{ secrets.GHCR_PULL_TOKEN }}
             project: ${{ vars.COMPUTESPHERE_PROJECT_ID }}
             environment: ${{ vars.COMPUTESPHERE_ENV_ID }}
         - run: echo "${{ steps.deploy.outputs.status }} → ${{ steps.deploy.outputs.service-url }}"
   

   You should seeA Deploy run starts in the Actions tab.

6. Watch it go live

   Open the running Deploy workflow. The build takes a minute or two, then the deploy step waits for Running. Open the URL it prints.

   Check yourself

   Every commit gets its own image tag. What does that give you?

   Each version maps to one commitFaster buildsA public image

   You should seeA green run whose last step prints created and your URL.

7. Change something

   Edit public/home.html, change some visible text, and commit to main. The old version serves until the new one is healthy; reload once the run is green.

   You should seeThe last step says updated, with the same URL, and the page shows your change.