fix(dispatcher): do not lose abort requests between progress controllers - #42247
fix(dispatcher): do not lose abort requests between progress controllers#42247Pavel Feldman (pavelfeldman) wants to merge 2 commits into
Conversation
An abort that arrives while the before/after instrumentation controller is running was dropped, because the controller handling the actual call was created afterwards. Track the pending abort error per call and pass it to every controller created for that call.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
Test results for "tests 1"4 flaky51121 passed, 1208 skipped Merge workflow run. |
| createProgressController(callMetadata: CallMetadata): ProgressController { | ||
| return ProgressController.createForSdkObject(this._object, callMetadata); | ||
| createProgressController(callMetadata: CallMetadata, pendingAbortError?: Error): ProgressController { | ||
| return ProgressController.createForSdkObject(this._object, callMetadata, pendingAbortError); |
There was a problem hiding this comment.
why can't it just throw here if there is a pending error?
Test results for "MCP"1 failed 8100 passed, 1311 skipped Merge workflow run. |
|
Hi, I'm the Playwright bot and I took a first look at the CI failures here. 🟢 The one failure is a pre-existing flake — this PR looks clear
DetailsThe latest reports show no real failures in the library suites (only flaky/interrupted), and a single MCP failure. Your change is confined to Pre-existing flake / infra
Flaky (rescued on retry), not caused by this PR
I'm a first pass, not the final word — if the annotate failure recurs on reruns of this exact commit, it's worth a second look, but right now nothing points at this PR. Triaged by the Playwright bot - agent run |
2 participants
Summary