test: add regression test for issue #1373 — default headersTimeout 300s

[SNTL84]

What

Adds a regression test test/issue-1373-default-timeout.js that confirms undici's default headersTimeout is 300e3 ms (300s), not the old 30s that caused issues in #1373.

Why

Issue #1373 was raised because the default fetch timeout was 30s, causing failures for slow-responding servers. The fix (aligning with Chrome's 300s default) appears to have been applied in:

• types/client.d.ts — Default: 300e3 ms
• docs/docs/api/Client.md — bodyTimeout default 300e3

However there was no regression test to prevent this from silently reverting. This PR fills that gap.

Test Approach

The test spins up an HTTP server that intentionally delays its response by 35 seconds — beyond the old 30s default but well within the 300s default. A HeadersTimeoutError would signal a regression.

test/issue-1373-default-timeout.js

Checklist

• Regression test added
• Follows existing test conventions (node:test, node:http, require('..'))
• No changes to source or docs needed (defaults are already correct)

---

Contributed by Milan Soni (SNTL84) · github.com/SNTL84
Open to OSS sponsorship & consulting — AI workflows, automation & full-stack builds · desidevloper.com

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.46%. Comparing base (907bc33) to head (c29b0b0).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #5467   +/-   ##
=======================================
  Coverage   93.46%   93.46%           
=======================================
  Files         110      110           
  Lines       37115    37115           
=======================================
  Hits        34690    34690           
  Misses       2425     2425           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[SNTL84]

All CI checks pass and coverage is stable at 93.46%. Would appreciate a maintainer review when you get a chance. 🙏

[mcollina]

The reason why this was never added is because I'm not waiting 30s for a test to pass.

Please use fake timers to implement this.

[SNTL84]

Thanks @mcollina — refactored to use mock.timers as suggested. The fake clock now ticks past the old 30s default (to 31s) instantly — no real delay, CI completes in milliseconds.

Changes made:

• Replaced real setTimeout(35_000) with mock.timers.enable() + mock.timers.tick(31_000)
• Switched from fetch() to Client for lower-level, more testable request handling
• Added proper t.after() cleanup hooks for both server and client

Please re-review when you get a chance. 🙏

— Milan Soni (SNTL84) · desidevloper.com

[SNTL84]

Hi @mcollina — I've addressed your feedback by switching to mock.timers as suggested. @metcoder95 has since approved. Could you take a quick look and dismiss your earlier review if you're satisfied? Thank you! 🙏