Repository navigation
Conversation
Member
Author
|
bugbot run |
Contributor
size-limit report 📦Compared: 5dab66a (PR test merge using develop at eb0e92e) Against: eb0e92e (develop) · base run #65112 ✅ Baseline matches: this diff contains only this PR's changes.
|
Contributor
Optional E2E testsSome optional checks were not successful. They do not block the merge. 6 not successful, 50 successful. See the run.
Successful checks
|
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
Member
Author
|
bugbot run |
Member
Author
|
bugbot run |
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
Member
Author
|
bugbot run |
Member
Author
|
bugbot run |
Nitro loads TanStack Start's server entry on the first request, so an instrument import at the top of the server entry only initializes the SDK during that request. Without node --import, as on Vercel, the first request of every server instance has no http.server span. The Vite plugin now adds a Nitro plugin that imports the server entry at startup. It imports the same chunk Nitro loads lazily, so the SDK is not bundled a second time. Fixes #25970 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…nit test Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Read the service entries from Nitro's vite-services module instead of deriving them, so the preload uses the same specifiers and skips server environments that are not Nitro services. Add a preloadServerEntry option to opt out, and check in the e2e test that the preload did not time out. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…o services Also formats the e2e Vite config. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
chargome
force-pushed
the
fix-tanstack-nitro-first
branch
from
November 4, 2026 14:11
0aad0ec to
7c08218
Compare
…e Nitro build Older Nitro betas generate the module that imports the services as #nitro-vite-setup, so the preload did nothing there. The preload module now loads that module itself, under either name. The service build leaves some packages as plain imports, so it could not see when one of them imports the Nitro runtime. Whether a service entry imports the preload module is now decided from the module graph of the Nitro build. Also logs with debug enabled, shares the Nitro options type, and no longer starts an unused app server in the init-in-server-entry e2e variant. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Member
Author
|
bugbot run |
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 3e30f13. Configure here.
| 'if (awaited.length > 0) {', | ||
| ' let loaded = false;', | ||
| ' let timeout;', | ||
| ` await Promise.race([Promise.all(awaited).then(() => { loaded = true; }), new Promise(resolve => { timeout = setTimeout(resolve, ${PRELOAD_TIMEOUT_MS}); })]);`, |
There was a problem hiding this comment.
Startup timer is not unref'd
Low Severity
The generated preload uses setTimeout without unref() or safeUnref(). That timer is injected into the Nitro server bundle, so a pending 3s fallback can keep a Node process alive. This was flagged because the project review rules require unref on timers in server runtime paths.
Triggered by project rule: PR Review Guidelines for Cursor Bot
Reviewed by Cursor Bugbot for commit 3e30f13. Configure here.
Member
Author
There was a problem hiding this comment.
Keeping the timer referenced on purpose. It only matters when the preload never settles, which happens with an import cycle the build did not detect. Then nothing else keeps the event loop alive, so with unref() Node would exit with code 13 (unsettled top-level await) before the timeout fires, which is the case the timeout exists for. The timer is cleared as soon as the preload finishes, and it keeps the process alive for at most 3s while the server starts.
Member
Author
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why: Nitro imports the TanStack Start server entry lazily on the first request. When
Sentry.initruns from aninstrument.serverimport in the server entry, as on Vercel where--importisn't available, it initializes too late for the first request'shttp.serverspan. The Vite plugin now adds a Nitro plugin that imports the service entries at server startup. It uses the same specifiers as Nitro, so no second SDK copy is bundled. The plugin uses top-level await so serverless platforms don't send the first request before init. Services that import the Nitro runtime are not awaited, to avoid an import cycle deadlock, and a 3s timeout is a fallback. You can turn it off withpreloadServerEntry: false.Unsure about: The preload reads Nitro's internal
#nitro/virtual/vite-servicestemplate. If that template changes shape, nothing is preloaded and we're back to the current behavior.Closes #25970