Worker: better reporting on socket errors - #1505
Merged
Merged
Conversation
Each run already opened a sentry isolation scope, but engine and socket callbacks fire outside its async context, so breadcrumbs and errors were landing on the global scope and interleaving across concurrent runs. The scope is now held on the run context and re-entered where the reporting actually happens. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
josephjclark
force-pushed
the
release/next
branch
from
August 26, 2026 14:21
06841f1 to
a14fce7
Compare
josephjclark
added a commit
that referenced
this pull request
Aug 27, 2026
* Worker: better reporting on socket errors (#1505) * Worker: attribute sentry reports to the run that produced them Each run already opened a sentry isolation scope, but engine and socket callbacks fire outside its async context, so breadcrumbs and errors were landing on the global scope and interleaving across concurrent runs. The scope is now held on the run context and re-entered where the reporting actually happens. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * carry payload size on step-complete events for better reporting * type fix * add payload size to step complete event * report payload size to sentry * changeset --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> * Worker: Fix dataclip serialization (#1507) * Worker: attribute sentry reports to the run that produced them Each run already opened a sentry isolation scope, but engine and socket callbacks fire outside its async context, so breadcrumbs and errors were landing on the global scope and interleaving across concurrent runs. The scope is now held on the run context and re-entered where the reporting actually happens. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * carry payload size on step-complete events for better reporting * type fix * report better errors out of sentry * Gate double-encoding fix for output dataclips behind WORKER_NO_STRINGIFY_STATE The worker double-JSON-encodes step output dataclips before sending them to Lightning: once via a manual stringify, then again when phoenix serializes the envelope. That double-encoding is what bloats large dataclips past Lightning's websocket frame limit and kills the connection mid-run. The runtime already sanitizes/clones state before it reaches this point, so the manual stringify is redundant - but skipping it changes the wire format, so it's opt-in via --stringify-state/--no-stringify-state (CLI) or WORKER_NO_STRINGIFY_STATE (env), default false to preserve current behaviour until the matching Lightning-side support (2.19+) is widely deployed. * docs and changelog --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> * versions --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
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
2 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.
This PR adds a few fixes to improve diagnostics
This so far won't fix any of the problems we've been seeing in #1504. I might include a fix.
AI Usage
Please disclose whether you've used AI anywhere in this PR (it's cool, we just
want to know!):
You can read more details in our
Responsible AI Policy