Harden Website Deploy workflow against transient GitHub Pages failures

[Copilot]

The Website Deploy job was failing intermittently in actions/deploy-pages with Deployment failed, try again later, despite a successful website build and uploaded Pages artifact. This change makes deployment resilient to transient Pages API failures without masking persistent failures.

• Deploy retry behavior

  • Keep the first actions/deploy-pages attempt as primary.
  • Add a second deploy attempt only when the first attempt fails.

• Failure semantics

  • If both attempts fail, fail the job explicitly with a clear error message.
  • This preserves signal for real deployment outages while reducing one-off flakes.

• Environment URL output handling

  • Resolve environment.url from whichever deploy attempt succeeds.

environment:
  name: github-pages
  url: ${{ steps.deployment.outputs.page_url || steps.deployment-retry.outputs.page_url }}

steps:
  - id: deployment
    continue-on-error: true
    uses: actions/deploy-pages@v5

  - if: steps.deployment.outcome == 'failure'
    id: deployment-retry
    continue-on-error: true
    uses: actions/deploy-pages@v5

  - if: steps.deployment.outcome == 'failure' && steps.deployment-retry.outcome == 'failure'
    run: |
      echo "Both website deployment attempts failed."
      exit 1

[mnonnenmacher]

@tsteenbe Can we close this? I saw only one failed deployment in the recent weeks due to a GitHub timeout and in this case it's unclear if a retry would even help.

[sschuberth]

@tsteenbe Can we close this? I saw only one failed deployment in the recent weeks due to a GitHub timeout and in this case it's unclear if a retry would even help.

Ping again @tsteenbe...