redis - test: Stop the keep-alive test from racing the connect timeout

[jaredwray]

Please check if the PR fulfills these requirements

• Followed the Contributing and Code of Conduct guidelines.
• Tests for the changes have been added (for bug fixes/features) with 100% code coverage. — test-only change

What kind of change does this PR introduce?

Test fix (flaky test). No source changes.

Problem

getClient > should keep a successful connection alive after connectionTimeout elapses connected to a real Redis with connectionTimeout: 50. The TCP connect and the handshake therefore had to beat a 50 ms timer. When every package's tests run in parallel in CI, they sometimes don't:

Error: Redis client is not connected or has failed to connect. This is thrown because throwOnConnectError is set to true.
 ❯ KeyvRedis.connectClient src/index.ts:440:11
 ❯ test/get-client.test.ts:705:18
Caused by: Error: Redis timed out after 50ms

It has failed this way twice today: in the codecov job on main at 6d4b270, and on the release PR #2192. The release workflow's test job runs the same suite, so this test can also block a release.

Fix

A bigger budget only makes a failure less likely. The first commit used 500 ms, and through a proxy that adds 250 ms to every chunk it still fails with Redis timed out after 500ms. Following #2182, the test now controls the clock it was racing instead of racing it:

• Keyv times the connect with a JS setTimeout (raceWithTimeout). The test fakes setTimeout and clearTimeout while it connects, so the race can't time out however long the handshake takes.
• The test then moves the clock past the timeout instead of sleeping. It runs in about 50 ms, down from 120 ms before and 600 ms with the first commit.
• node-redis's TCP connectTimeout runs on Node's internal timers, which stay real. connectionTimeout is therefore 1 s, the window mono - test: Stop expiry tests from racing a 100 ms TTL #2182 gave the tests that use the server's clock. It only bounds the TCP connect, not the handshake.
• node-redis only uses the global setTimeout for its ping timer, which is off unless pingInterval is set. Real timers come back before set/get, and a finally restores them and disconnects even if an assertion fails.

The test also checks that the same client is still open after the timeout. Before, a client torn down at the timeout went unnoticed, because the adapter reconnects on the next call. No retries were added (see AGENTS.md).

Verification

Scenario	Original (50 ms)	First commit (500 ms)	This PR
Direct Redis	pass	pass	pass, 10/10 runs
Proxy adding 40 ms per chunk	fails (timed out after 50ms, the CI error)	pass	pass
Proxy adding 250 ms per chunk (handshake over 1 s)	fails	fails (timed out after 500ms)	pass
Simulated bug: a timer destroys the client once the timeout elapses	passes (misses it)	passes (misses it)	fails (client.isOpen is false)

• The whole get-client.test.ts file passes (35/35).
• biome check --error-on-warnings is clean.
• tsc on the test file shows no new errors (0 before, 0 after).

---

🤖 Generated with Claude Code

https://claude.ai/code/session_01Cwt8ciN4uKrQkEM14TCff2

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-04T20:22:26.526815Z	ea48518	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (d0bc310) to head (59a9b5d).

Additional details and impacted files

@@            Coverage Diff            @@
##              main     #2193   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           56        56           
  Lines         5802      5802           
  Branches       996       988    -8     
=========================================
  Hits          5802      5802           

☔ 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.