meta(changelog): Update changelog for 10.33.0 #18770
Merged
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
[Gitflow] Merge master into develop
Adds 3 size-limit entries for logs, metrics and both combined. I have some optimizations lined up for reducing bundle size in both but I need a baseline to see if they actually do any good.
…ners (#18581) Previously, we called `onHidden` in our slightly modified version of the vendored in `whenIdleOrHidden` API from web-vitals. This caused 2 additional event listeners to be registered with each `whenIdleOrHidden` invocation, which also didn't get cleaned up properly. To fix this, and prevent similar situations, this PR: - inlines the `pagehide` event listener registration in `whenIdleOrHidden` so that we can remove the `onHidden` call - adds `once: true` to the listener registration so that the callback is only invoked once - deprecates `onHidden` because IMHO we should remove it in v11 and replace the one remaining use of it with a direct `visibilityChange` subscription. Closes #18584 (added automatically) closes https://linear.app/getsentry/issue/JS-1339/investigate-multiple-event-listeners-in-nextjs-sdk
So far if multiple expected envelopes were chained on the cloudflare test runner, the runner expected them to arrive in the order pre-defined by the caller, else the test would fail. This can sometimes be too strict, resulting in flakes. This PR adds an unordered mode to the runner that lifts the order restriction on incoming envelopes. This mode is also used in the hono cloudflare test now. Closes #18589
…18545) (closes #18534) (closes [JS-1319](https://linear.app/getsentry/issue/JS-1319/cloudflare-transport-doesnt-consume-fetch-response-bodies)) This consumes the body so it is safe to be closed by Cloudflare Workers. Unfortunately this cannot be reproduced locally and is only happening when the worker is being deployed (so there is no E2E test for this) @AbhiPrasad do you think it is necessary to add this code snippet in the other transports we have? It is what I've read specifically to Cloudflare, since they seem to be super strict on it, but this could potentially be a problem in other runtimes as well.
Pino allows using custom namespaces and this PR makes sure they are actually used. The symbols are extracted with a method, although Pino would allow getting them from `pino.symbols` ([docs](https://getpino.io/#/docs/api?id=pino-symbols)). However, we only have access to the `logger` instance. Closes #18594
E2E tests for solid/vue tanstack router fail starting `1.42.x`. Pinning these for now until we have figured out what's going on. Closes #18612 (added automatically)
https://github.com/cloudflare/agents/releases/tag/agents%400.2.34 started externalizing the ai package. Pinning this to the version before this let's our tests pass. When bumping the `agents` package to `0.3.0` and adding `ai` with v6 types start breaking. Closes #18611 (added automatically)
The example pages the Sentry wizard adds to the codebase use `.diagnoseSdkConnectivity()`. It's also possible to already set up a tunnel through the wizard. However, this tunnel was previously not respected. Closes #17991
Ref / Resolves: #18089 `hoist-non-react-statics` package we depend on is not ESM compatible, and causing issues for users on Cloudflare environments. As this package has not been updated in 5 years, and our use has not changed for long time, I think we can just inline this to make it compatible, and save users from explicitly externalizing this package. --------- Co-authored-by: Charly Gomez <charly.gomez@sentry.io>
### Summary Removes `resolve` dependency and replaces it with `createRequire` which would work in esm modules as well, since our min Node.js version is 18 then it should be fine to do this replacement. This should reduce the npm install size and the SSR bundle size for the SDK. closes #12860
This came up while working on #18580 The span ops should be using snake_case as per our spec: https://develop.sentry.dev/sdk/telemetry/traces/span-operations/
…18620) When using [exactOptionalPropertyTypes](https://www.typescriptlang.org/tsconfig/#exactOptionalPropertyTypes), it is pretty hard to match Fastify's instance types given all the missing overloads from the partial signature we have. The problematic area is around `addHook`. We either need to have all the signatures implemented exactly, or outright import Fastify's types which is not ideal since it is not a dependency. I opted to relax the types specifically for `setupFastifyErrorHandler` by providing a minimal instance with the one method it needs to avoid TS trying to matching the other properties and methods signatures, not ideal but simple enough. I verified that this works for Fastify v3 throughout v5 closes #18619
…#18628) The `offline.test.ts` unit tests in core have been bugging me for a long time because they take almost 30s to run locally. Which makes running tests in watch mode in core quite annoying. The core problem in these tests was that we used real timers and waited for up to 25s in one test. This PR: - uses vitest's fake timers instead of real timers - removes the `waitUntil` calls in favor of advancing timers. Mostly by the upper bound we specified before, though in some situations, I had to use a lower value because the previous waitUntil condition was satisfied earlier. - re-enables the rate limiting test which was flaky before. I'm curious if it'll flake again but my guess is that with the more strict fake timers, it might just behave as we hope. Now, the tests only take a couple of ms to complete. Closes #18630 (added automatically)
Setting an IPv6 URL as a DSN previously didn't work, see #2996. This patch updates the `DSN_REGEX` to correctly match IPv6 and fixes the then occuring issue that the brackets "[" and "]" are in the request's hostname and prevent the request from being made. --------- Co-authored-by: Lukas Stracke <lukas.stracke@sentry.io>
This PR adds the external contributor to the CHANGELOG.md file, so that they are credited for their contribution. See #17708 Co-authored-by: Lms24 <8420481+Lms24@users.noreply.github.com>
…tion (#18416) Adds the previously introduced `isolateTrace` option to the `node-cron` instrumentation. This ensures that independent tasks triggered by the same cron have distinct traces when `isolateTrace` is set.
…edback (#18492) 1. fixed incorrect Bun version command 2. removed unused variable 3. improved user feedback(logs a clear message when Bun is already on the latest version instead of silently exiting) --------- Co-authored-by: Lukas Stracke <lukas.stracke@sentry.io>
This PR adds the external contributor to the CHANGELOG.md file, so that they are credited for their contribution. See #18416 Co-authored-by: Lms24 <8420481+Lms24@users.noreply.github.com>
This PR adds the external contributor to the CHANGELOG.md file, so that they are credited for their contribution. See #18492 Co-authored-by: Lukas Stracke <lukas.stracke@sentry.io>
…orFilterIntegration` (#18632) Some users get flooded by third party errors despite having set their third party error filtering to `"drop-error-if-contains-third-party-frames"`. This comes from the fact that often the stacktrace includes our internal wrapper logic as last trace in the stack. With this PR we're trying to work around this by specifically ignoring these frames with a new opt-in mechanism. Marked this as experimental, so users will know this option might lead to errors being misclassified. - Adds a new experimental option `experimentalExcludeSentryInternalFrames` to the `thirdPartyErrorFilterIntegration` - Once enabled we apply a strict filter for frames to detect our internal wrapping logic and filter them out so they do not misclassify injected code as internal errors. Closes #13835
…gs (#18638) With this PR we emit a warning for client, node and edge whenever the user sets `debug: true` in `init` but at the same time treeshakes logging statements using webpack. This will not emit anything for turbopack. closes https://linear.app/getsentry/issue/FE-610/warn-at-build-time-for-conflicting-disablelogger-and-debug-settings-in
Ensures Vercel AI span op is always set correctly at span start, regardless of whether the model ID is available. Changes: - Removed the model ID check gate before calling processGenerateSpan - the op is determined by span name, not model ID - Span name updates (e.g., generate_text gpt-4) are now only applied when model ID exists, avoiding undefined in names - Added integration test for late model ID scenario Closes #18448
This PR adds the external contributor to the CHANGELOG.md file, so that they are credited for their contribution. See #18024 Co-authored-by: Lms24 <8420481+Lms24@users.noreply.github.com>
…ons (#17583) This adds a better way to ensure we flush in node-based vercel functions. Note that this _does not_ work in edge functions, so we need to keep the `vercelWaitUntil` code around for this - but we can noop it in node and instead rely on the better way to handle this, I believe. @chargome when you're back, can you verify if this makes sense? 😅 Closes #17567 --------- Co-authored-by: Charly Gomez <charly.gomez@sentry.io>
Pre-work for #18160 We have multiple places in which we need to combine the `ScopeData` of global, isolation and current scopes, to apply this data to telemetry items (events, logs, metrics, soon also spansV2). Previously, we did this in-place or with helpers that were not re-used. This PR now unifies the various locations to one helper from core which can be reused everywhere. Closes #18586 (added automatically)
…#18756) The `thirdPartyErrorFilterIntegration` was not able to identify first-party worker code as because module metadata stayed in the worker's separate global scope and wasn't accessible to the main thread. We now forward the metadata the same way we forward debug ids to the main thread which allows first-party worker code to be identified as such. Closes: #18705
Applies scope attributes to metrics, analogously to how we apply them to logs (see #18184). Added unit and integration tests.
Currently the cache components feature in Next.js prevents us from using any random value APIs like: - `Date.now` - `performance.now` - `Math.random` - `crypto.*` We tried resolving this by patching several span methods, but then we have plenty of other instances where we use those APIs, like in trace propagation, timestamp generation for logs, and more. Running around and patching them one by one in the Next.js SDK isn't a viable solution since most of those functionalities are strictly internal and cannot be patched from the outside, and adding escape hatches for each of them is not maintainable. So I'm testing out the other way around, by hunting those APIs down and wrapping them with a safe runner that acts as an escape hatch. Some of the Vercel engineers suggested doing that, but we need to do it for ~almost every call~ (see Josh [comment](#18700 (comment)) below). The idea is an SDK can "turn on" the safe runner by injecting a global function that executes a callback and returns its results. I ### How does this fix it for Next.js? The Next.js SDK case, a safe runner would be an `AsyncLocalStorage` snapshot which is captured at the server runtime init, way before any rendering is done. ```ts const sym = Symbol.for('__SENTRY_SAFE_RANDOM_ID_WRAPPER__'); const globalWithSymbol: typeof GLOBAL_OBJ & { [sym]?: SafeRandomContextRunner } = GLOBAL_OBJ; globalWithSymbol[sym] = AsyncLocalStorage.snapshot(); // core SDK then offers a fn to run any random gen function export function withRandomSafeContext<T>(cb: () => T): T { // Looks for the global symbol and if it is set it uses the runner // otherwise just runs the callback normally. } ``` I kept the API internal as much as possible to avoid users messing up with it, but the `@sentry/opentelemetry` SDK also needed this functionality so I exported the API with `_INTERNAL` prefix as we already do. --- I tested this in a simple Next.js app and it no longer errors out, and all current tests pass. I still need to take a look at the traces and see how would they look in cached component cases. Charly is already working on this and may have a proper solution, but I thought to just see if we can ship a stopgap until then. On the bright side, this seems to fix it as well for Webpack. closes #18392 closes #18340
…#18762) Adds support for applying an application key to WASM stack frames that can be then used in the `thirdPartyErrorFilterIntegration` for detection of first-party code. This changes how `thirdPartyErrorFilterIntegration` deals with native frames to also check for a `instruction_addr` for WASM-native code. Usage: ```js Sentry.init({ integrations: [ wasmIntegration({ applicationKey: 'your-custom-application-key' }), ←───┐ thirdPartyErrorFilterIntegration({ │ behaviour: 'drop-error-if-exclusively-contains-third-party-frames', ├─ matching keys filterKeys: ['your-custom-application-key'] ←─────────────────────────┘ }), ], }); ``` Closes: #18705
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
dev-packages/browser-integration-tests/suites/public-api/metrics/simple/test.ts
Show resolved
Hide resolved
Lms24
force-pushed
the
prepare-release/10.33.0
branch
from
73b1f80 to
10fa35e
Compare
Contributor
size-limit report 📦
|
Contributor
node-overhead report 🧳Note: This is a synthetic benchmark with a minimal express app and does not necessarily reflect the real-world performance impact in an application.
|
dev-packages/node-core-integration-tests/suites/ipv6/scenario.ts
Outdated
Show resolved
Hide resolved
Member
Author
|
Ready |
I accidentally left a hard-coded SDK version in a test recently added by me. This PR fixes that so that CI doesn't break on the next release Closes #18772 (added automatically)
Lms24
force-pushed
the
prepare-release/10.33.0
branch
from
10fa35e to
99d5597
Compare
chargome
approved these changes
Jan 12, 2026
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Lms24
force-pushed
the
prepare-release/10.33.0
branch
from
99d5597 to
1630389
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
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.
No description provided.