♻️ replace custom cookie lock with native Web Locks API

[thomas-lebeau]

Motivation

The old session concurrency mechanism was a custom cookie-based spin-lock — polling for a lock= cookie entry with retries and delays. It was complex (~136 lines in sessionStoreOperations.ts), only enabled in Chromium (isLockEnabled: isChromium()), and left Firefox/Safari without any cross-tab session protection.

Changes

• Remove sessionStoreOperations.ts and its cookie-based lock mechanism
• Introduce sessionLock.ts using navigator.locks for concurrency — available in all modern browsers (Chrome 69+, Firefox 96+, Safari 15.4+), so this also extends cross-tab session consistency to non-Chromium browsers
• Simplify sessionStore.ts by inlining retrieve/persist logic directly within the lock callback
• Remove isLockEnabled from SessionStoreStrategy interface
• Clean up related tests and fake session store strategy

Net: ~554 lines removed, ~137 added.

Test instructions

Unit tests cover the async initialization scenarios:

• Session manager initialization is delayed until the lock is released
• onReady callback is deferred until after the lock is released
• Session is expired when tracking consent is withdrawn while waiting for the lock

Checklist

• Tested locally
• Tested on staging
• Added unit tests for this change.
• Added e2e/integration tests for this change.
• Updated documentation and/or relevant AGENTS.md file

[cit-pr-commenter-54b7da]

Bundles Sizes Evolution

📦 Bundle Name	Base Size	Local Size	𝚫	𝚫%	Status
Rum	170.70 KiB	169.79 KiB	-932 B	-0.53%	✅
Rum Profiler	6.18 KiB	6.18 KiB	-1 B	-0.02%	✅
Rum Recorder	25.26 KiB	25.48 KiB	+226 B	+0.87%	✅
Logs	56.94 KiB	56.17 KiB	-787 B	-1.35%	✅
Flagging	944 B	944 B	0 B	0.00%	✅
Rum Slim	127.92 KiB	127.17 KiB	-769 B	-0.59%	✅
Worker	23.63 KiB	23.63 KiB	0 B	0.00%	✅

🚀 CPU Performance

Action Name	Base CPU Time (ms)	Local CPU Time (ms)	𝚫%
RUM - add global context	0.0047	0.0045	-4.26%
RUM - add action	0.0136	0.0145	+6.62%
RUM - add error	0.0141	0.0154	+9.22%
RUM - add timing	0.003	0.0039	+30.00%
RUM - start view	0.0144	0.0127	-11.81%
RUM - start/stop session replay recording	0.0007	0.0007	0.00%
Logs - log message	0.0161	0.0171	+6.21%

🧠 Memory Performance

Action Name	Base Memory Consumption	Local Memory Consumption	𝚫
RUM - add global context	26.87 KiB	31.31 KiB	+4.44 KiB
RUM - add action	49.71 KiB	222.03 KiB	+172.32 KiB
RUM - add timing	26.39 KiB	34.25 KiB	+7.87 KiB
RUM - add error	56.25 KiB	245.16 KiB	+188.90 KiB
RUM - start/stop session replay recording	25.90 KiB	33.96 KiB	+8.06 KiB
RUM - start view	451.66 KiB	574.31 KiB	+122.65 KiB
Logs - log message	45.22 KiB	223.29 KiB	+178.06 KiB

🔗 RealWorld

[datadog-official]

✅ Tests

🎉 All green!

❄️ No new flaky tests detected
🧪 All tests passed

🎯 Code Coverage (details)
• Patch Coverage: 75.51%
• Overall Coverage: 76.72% (+0.03%)

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: 2ee6b36 | Docs | Datadog PR Page | Was this helpful? React with 👍/👎 or give us feedback!}

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: e7defdcc66

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".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[mormubis]

All unit tests replace withNativeSessionLock with a synchronous stub — is the real navigator.locks async path only exercised in E2E? If so, a timing bug there would not be caught by unit tests.