← Recipes

阅读中文版本

Use HTTP/2 Server Push for Critical Resources

Push the CSS and font before the browser parses the HTML so the waterfall collapses and LCP drops by 400ms.

~20 minperformancehttp2server-push

Task

Apply Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font to reduce a measured bottleneck that shows up in the performance budget or Lighthouse audit.

When to use this

Lighthouse reports the page fails the performance budget because of Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font, scoring 72 on mobile.

Prerequisites

  • A Lighthouse or WebPageTest baseline measurement
  • Access to the build configuration

Steps

1. Configure Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font in the project configuration file with the recommended defaults.

Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font

Expected output: The configuration is accepted and Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font runs without errors on the first invocation.

2. Run the check locally to verify Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font produces the expected result before pushing.

pnpm test

Expected output: Local run matches CI behavior; no surprises after push.

3. Add the gate to the CI pipeline so Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font blocks merge on failure.

- name: Run add http2 server push
  run: pnpm test

Expected output: Pull requests with failing Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font are blocked from merging.

Troubleshooting

Problem

The check passes locally but fails in CI

Solution

Pin dependency versions and the runtime to match CI exactly; reproduce with the same Docker image or act.

Problem

Configuration is rejected with a schema error

Solution

Check the tool version matches the config schema; older versions may not support newer fields.

Limitations

  • The gate only runs on events that trigger the workflow; locally pushed commits bypass it.
  • Configuration drift is possible if someone edits settings outside version control.

Measure before optimizing

Start by measuring the problem. Push the CSS and font before the browser parses the HTML so the waterfall collapses and LCP drops by 400ms. The most common mistake is adding the gate without first confirming it catches the issue you care about. Run Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font against the current codebase and inspect the output before wiring it into CI.

Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font

Verify the improvement

Once the gate is in place, make failures actionable. A failing Link: </style.css>; rel=preload; as=style, </font.woff2>; rel=preload; as=font should point to the exact file and line, not a generic error. Pin the tool version in the lockfile so a upstream release does not silently change behavior. Review the gate output in every pull request just like a test result.