Learn how to share artifacts between jobs in GitHub CI/CD workflows

Learn how to share artifacts between jobs in GitHub CI/CD workflows

In GitHub Actions, artifacts allow files or directories to be shared between jobs in a workflow. This can be useful for passing data, build outputs, test results, or deployment artifacts across different steps or jobs.

1. Creating and Uploading Artifacts

Upload-artifact Action

The actions/upload-artifact action allows files to be uploaded as artifacts. These artifacts can then be downloaded by subsequent jobs in the workflow.

Example: Uploading Artifacts

xxxxxxxxxx
24
1
jobs:
2
  build:
3
    runs-on: ubuntu-latest
4
    steps:
5
     - name: Checkout Repository
6
        uses: actions/checkout@v3
7
     - name: Build Artifacts
8
        run: ./build.sh
9
        uses: actions/upload-artifact@v3
10
        with:
11
          name: build-artifacts
12
          path: build/
13
  deploy:
14
    needs: build
15
    runs-on: ubuntu-latest
16
    steps:
17
     - name: Checkout Repository
18
        uses: actions/checkout@v3
19
     - name: Download Artifacts
20
        uses: actions/download-artifact@v3
21
        with:
22
          name: build-artifacts
23
     - name: Deploy
24
        run: ./deploy.sh --build ./build/

2. Downloading Artifacts

Download-artifact Action

The actions/download-artifact action retrieves previously uploaded artifacts and makes them available for use in subsequent jobs.

Example: Downloading Artifacts

xxxxxxxxxx
24
1
jobs:
2
  build:
3
    runs-on: ubuntu-latest
4
    steps:
5
     - name: Checkout Repository
6
        uses: actions/checkout@v3
7
     - name: Build Artifacts
8
        run: ./build.sh
9
        uses: actions/upload-artifact@v3
10
        with:
11
          name: build-artifacts
12
          path: build/
13
  deploy:
14
    needs: build
15
    runs-on: ubuntu-latest
16
    steps:
17
     - name: Checkout Repository
18
        uses: actions/checkout@v3
19
     - name: Download Artifacts
20
        uses: actions/download-artifact@v3
21
        with:
22
          name: build-artifacts
23
     - name: Deploy
24
        run: ./deploy.sh --build ./build/

3. Artifact Retention

GitHub provides configurable artifact retention to manage how long artifacts are stored after a workflow completes.

Setting Artifact Retention

You can specify how long artifacts should be retained:

xxxxxxxxxx
29
1
jobs:
2
  deploy:
3
    needs: build
4
    runs-on: ubuntu-latest
5
    outputs:
6
      build-artifact: ${{ steps.download-artifact.outputs.artifact_path }}
7
    steps:
8
     - name: Download Artifacts
9
        uses: actions/download-artifact@v3
10
        with:
11
          name: build-artifacts
12
     - name: Deploy
13
        run: ./deploy.sh --build ${{ steps.download-artifact.outputs.artifact_path }}
14
    artifact:
15
      name: build-artifacts
16
      paths: 
17
       - build/
18
    outputs:
19
      artifact_path: ${{ steps.upload-artifact.outputs.artifact_path }}
20
  cleanup:
21
    if: always()
22
    runs-on: ubuntu-latest
23
    needs: deploy
24
    steps:
25
     - name: Clean Up Artifacts
26
        uses: actions/artifact-cleanup@v3
27
        with:
28
          name: build-artifacts
29
          retention-days: 7

In this example:

Artifacts are retained for 7 days before being deleted automatically.

4. Deleting Artifacts

You can manually delete artifacts via actions using the actions/artifact-cleanup action or automatically after a specified retention period.

Deleting Artifacts

xxxxxxxxxx
9
1
jobs:
2
  cleanup:
3
    runs-on: ubuntu-latest
4
    steps:
5
     - name: Clean Up Old Artifacts
6
        uses: actions/artifact-cleanup@v3
7
        with:
8
          name: build-artifacts
9
          retention-days: 30

This deletes all artifacts older than 30 days.

5. Using Permissions

• Upload Permissions: Grant necessary permissions for uploading artifacts:

  xxxxxxxxxx
  2
  1
  permissions:
  2
  artifacts: write
  

• Download Permissions: Ensure artifacts can be downloaded:

  xxxxxxxxxx
  2
  1
  permissions:
  2
  artifacts: read
  

6. Artifact Limits and Storage

GitHub imposes limits on artifact sizes and total storage:

• Maximum artifact size: 100 GB per workflow run.

• Total storage limit: 2 GB per user (across all repositories).

Best Practices for Using Artifacts

1. Granular Artifacts: Create multiple smaller artifacts for different types of outputs (e.g., build artifacts, test reports).

2. Limit Retention: Set a reasonable retention period for artifacts to manage storage costs and reduce the lifecycle of unused artifacts.

3. Secure Access: Ensure sensitive data is stored securely, especially when handling API keys or other sensitive information.

4. Error Handling: Implement error handling in workflows if artifact operations fail (e.g., downloads, deletions).

5. Matrix Strategy: Use artifacts effectively in matrix testing to share results between environments (e.g., different OS, Node.js versions).