[pull] master from getsentry:master #42
Open
Conversation
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
![pull[bot]](https://avatars.githubusercontent.com/in/12910?s=60&v=4){kind=small}
feat(react-router): Create a Vite plugin that injects sentryConfig into the global vite config
This updates naming etc. for express (v5) tests, and makes sure that we test this in esm & cjs in a basic way. It also adds test for handling of the root route (`/`) as well as 404, where current failures are shown (see #16203).
As pre-work for #16178, actually add a test for this (kind of). This showed that there was actually a fundamental flaw here, as we looked as the `req` not the `res`, oops.
This can be configured like this:
```js
httpIntegration({
dropSpansForIncomingRequestStatusCodes: [404, [300,399]]
})
```
It defaults to `[404]`.
Closes #16193
resolves #16110 Stores the Client to LogBuffer WeakMap onto the global object so logs can be retrieved correctly during flushing. Previously, the WeakMap reference would be different at flush time, causing no logs to be found for any given client. Adds an express e2e tests to ensure logs are flushed correctly.
This PR enables reading the `SENTRY_RELEASE` variable from the CF `env` that users should pass to their `withSentry` worker wrapper. This is quite similar to how we'd usually access env variables in Node-based SDKs. We need this for uploading release-based source maps for CF worker functions being bundled and deployed by wrangler. see getsentry/sentry-wizard#824 (comment) ref https://linear.app/getsentry/issue/WIZARD-36/improve-sentrywizard-i-sourcemaps-for-cloudflarewrangler --------- Co-authored-by: Abhijeet Prasad <aprasad@sentry.io>
Co-authored-by: Charly Gomez <charly.gomez@sentry.io>
meta(changelog): Update changelog for 9.16.0
[Gitflow] Merge master into develop
We could never get this to apply properly on vercel in production, so we're removing this for now and may revisit this later. Replaces #16178
meta(changelog): Update changelog for 9.16.1
[Gitflow] Merge master into develop
This deprecates the `ignoreEmberOnErrorWarning` option, as it is no longer used. Previously, we logged a warnining if we detected that this was used. But the option is now deprecated/will be removed, and since there is no real logic attached to this we can simply remove this (as well as the usage of the global `Ember` namespace) to avoid deprecations. Closes #16504 --------- Co-authored-by: Lukas Stracke <lukas.stracke@sentry.io>
[Gitflow] Merge master into develop
This changes the behavior of `suppressTracing` to be less problematic
(or, problematic in a different way).
Today, because in the browser we do not have async context isolation,
there is only a single shared top scope. Because of this, if you have
code like this:
```js
const spanPromise = suppressTracing(async () => {
await new Promise(resolve => setTimeout(resolve, 100));
return startInactiveSpan({ name: 'span' });
});
const span = startInactiveSpan({ name: 'span2' });
```
The span2 will also be suppressed, because `suppressTracing` forks the
scope and sets data on it to ensure spans are not recorded. This is
problematic as that will result in completely unrelated spans, e.g.
pageload/navigation spans, being suppressed while e.g. an ongoing fetch
call that is suppressed happens.
This PR changes this to instead only suppress tracing synchronously in
the browser. This obviously is also not really ideal and can lead to
things _not_ being suppressed, but it feels like the better tradeoff for
now.
By default, the instrumentation will register span processors only when
the ai package is used. This is done to avoid overhead of span
processing for users that do not even use this package.
However, it seems that in some environments, esp. in Next.js, the
instrumentation is not added correctly, thus never running this, and not
converting spans correctly. For now, this PR adds an escape hatch to
manually opt-in to this to still get correct spans:
```js
vercelAiIntegration({ force: true })
```
…16553) ref https://linear.app/getsentry/issue/FE-503/investigate-nested-middleware-spans-in-webfx-koa-application The Koa integration in `@sentry/node` was updated to expose the `ignoreLayersType` option from `@opentelemetry/instrumentation-koa`, aligning its configuration with the GraphQL integration. https://www.npmjs.com/package/@opentelemetry/instrumentation-koa <span><div class="markdown-heading"><h3 class="heading-element">Koa Instrumentation Options</h3><a id="user-content-koa-instrumentation-options" class="anchor" aria-label="Permalink: Koa Instrumentation Options" href="https://www.npmjs.com/package/@opentelemetry/instrumentation-koa#koa-instrumentation-options"></a></div></span><span> Options | Type | Example | Description -- | -- | -- | -- ignoreLayersType | KoaLayerType[] | ['middleware'] | Ignore layers of specified type. requestHook | KoaRequestCustomAttributeFunction | (span, info) => {} | Function for adding custom attributes to Koa middleware layers. Receives params: Span, KoaRequestInfo. <p><code>ignoreLayersType</code> accepts an array of <code>KoaLayerType</code> which can take the following string values:</p> <ul> <li> <code>router</code>,</li> <li> <code>middleware</code>.</li> </ul></span> --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com>
…16558) A new "Debugging Tests" section was added to `dev-packages/node-integration-tests/README.md`. This section documents the `DEBUG` environment variable, which enables verbose logging for the integration test suite. When `DEBUG=1` is set (e.g., `DEBUG=1 yarn test`), the test runner provides detailed output, including: * Test scenario startup information (path, flags, DSN). * Docker Compose output when tests use `withDockerCompose`. * Child process stdout and stderr. * HTTP requests made during tests. * Process errors and exceptions. * Line-by-line output from test scenarios. This addition improves discoverability and understanding of the debugging capabilities, aiding in troubleshooting failing tests and analyzing test execution flow. --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Bump vendored `web-vitals` library. Important changes from the original library: - For now until the next SDK major, we'll keep reporting FID. `web-vitals` removed the already deprecated APIs for it in v5 but we simply keep them from v4 - `web-vitals` further removed compatibility for older iOS Safari versions. Unfortunately, [we still support Safari 14.4 ](https://docs.sentry.io/platforms/javascript/troubleshooting/supported-browsers/) which is the last version that doesn't yet fully support the `visibilitychange` event. This requires us to keep the `onHidden` helper around which also listens to `pagehide` events that this Safari version supports. I adjusted our integration tests to keep one around that fails if we remove this special handling in a future upgrade (also added some context for future us). I will follow up with at least one more PR to do some more refactorings but I decided to keep them minimal in this PR to get better diffs for reviewing: - rename the `get*` files to `on*`, since this is how they're named now in the official library closes #16310
This PR fixes a bug discovered in #16486 (comment) where the root component span would not end correctly if `trackComponents` was `false`. Also added a comment to explain the purpose of the root component `ui.vue.render` span. We might want to look into renaming or removing this span in the future but for now, let's fix the behaviour.
1. Pins `minimatch` in nestjs-11 E2E test app -See nestjs/nest#15273. We can drop this when we bump node to 20.19.0, but for now this should be fine. 2. Pin trpc for nextjs-t3 E2E test app to 11.3.0 - See trpc/trpc#6821
meta(changelog): Update changelog for 9.29.0
…16566) Extracted out from #16565 I noticed that our `modulesIntegration` is pretty limited: 1. It does nothing on EMS 2. It does not work on Next.js (even though that is CJS) This PR makes this a bit more robust (not perfect): 1. It generally now also tries to look at `process.cwd() + 'package.json'` and take the dependencies and devDependencies from there. this should generally work in esm apps now as well, at least at a basic level. You do not get all dependencies and versions may be ranges, but better than nothing. 2. For Next.js, we inject a modules list based off the package.json at build time, as we do not have proper access to this at runtime.
[Gitflow] Merge master into develop
…le is detected (#16565) This PR improves the handling of the `ai` instrumentation to always be enabled when we detect the `ai` module is installed. For this, we leverage the `modulesIntegration`. This PR should make usage of ai instrumentation in Next.js "automatically" again - BUT users will have to specific ` experimental_telemetry: { isEnabled: true },` at each call manually for the time being. --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com>
…16557) resolves #16237 In #16348 we had to revert the PR that added `detail` to `measure` spans as attributes. measure API: https://developer.mozilla.org/en-US/docs/Web/API/Performance/measure detail: https://developer.mozilla.org/en-US/docs/Web/API/PerformanceMeasure/detail This was [reverted](#16347) because it was causing issues in firefox, specifically this error was being thrown ``` Error: Permission denied to access object at _addMeasureSpans(../../node_modules/@sentry-internal/browser-utils/build/esm/metrics/browserMetrics.js:255:41) at X2e/<(../../node_modules/@sentry-internal/browser-utils/build/esm/metrics/browserMetrics.js:194:9) at addPerformanceEntries(../../node_modules/@sentry-internal/browser-utils/build/esm/metrics/browserMetrics.js:174:48) at idleSpan.beforeSpanEnd(../../node_modules/@sentry/browser/build/npm/esm/tracing/browserTracingIntegration.js:90:9) at span.endapply(../../node_modules/@sentry/browser/node_modules/@sentry/core/build/esm/tracing/idleSpan.js:52:9) at Coe/<(../../node_modules/@sentry/browser/node_modules/@sentry/core/build/esm/tracing/idleSpan.js:196:12) at sentryWrapped(../../node_modules/@sentry/browser/build/npm/esm/helpers.js:38:17) ``` From debugging, this seems to be coming from a `DOMException` being thrown @https://developer.mozilla.org/en-US/docs/Web/API/DOMException This was re-implemented, and then we added tests to validate that this wouldn't break on firefox. --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com>
…#16590) Noticed this small but potentially impactful typo: We used to mark all finished spans as sent, not just the ones that are actually sent :O (Also we iterated twice over the same array, streamlined this by just doing this once).
…ctions issues (#16500) Next.js’s uses an internal AsyncLocalStorage (ALS) to share minimal request context—it deliberately does not populate the full Request object or URL in ALS, so only headers (Referer, x-forwarded-host, x-forwarded-proto) are available at runtime in Server Components. To still capture a usable URL for metadata we use a three-step best-effort fallback: 1. `Referer` header Use it as the fully qualified URL if present (note: this could be omitted by the browser). 2. Proxy headers + route Reconstruct protocol://host from x-forwarded-proto / x-forwarded-host (or host), then append the normalized component route with substituted [param] values. Query strings and fragments are TBD as query params fetched from args are always undefined for some reason, this needs investigation 3. Span fallback As a last resort, use the parent span’s http.target attribute (which contains the path and query) from the root HTTP span. <img width="1261" alt="Screenshot 2025-06-02 at 4 04 31 PM" src="https://github.com/user-attachments/assets/0cebe68a-89e7-4b61-9c58-962f8e70c7f9" /> closes: https://linear.app/getsentry/issue/JS-487/capture-requesturl-for-nextjs-server-spans
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.
See Commits and Changes for more details.
Created by
pull[bot]
Can you help keep this open source service alive? 💖 Please sponsor : )