aglinxinyuan commented on code in PR #5781:
URL: https://github.com/apache/texera/pull/5781#discussion_r3464052781
##########
amber/src/main/scala/org/apache/texera/web/service/WorkflowService.scala:
##########
@@ -292,11 +292,23 @@ class WorkflowService(
executionService.onNext(execution)
execution.executeWorkflow()
} catch {
- case e: Throwable => errorHandler(e)
+ case e: Throwable =>
+ errorHandler(e)
+ reportFatalErrorsToSubscribers(executionStateStore)
Review Comment:
Thanks — yes, exactly those two phases (pre-publish: report directly via
`errorSubject`; post-publish: route through the metadata-store diff handler).
On the alternative: I agree a single reporting site bundled with the
ExecutionService is cleaner. I'd prefer to keep it out of this focused fix,
though — making `WorkflowExecutionService` construction error-free and moving
the throwing work into a separate `init()` is a lifecycle refactor of the
service (its constructor currently makes the external calls that are exactly
the pre-publish failure window this PR surfaces, and several call sites rely on
construction side effects). Happy to take that on as a dedicated follow-up if
you'd like; for now this PR keeps the two-phase approach you confirmed.
_🤖 Addressed by [Claude Code](https://claude.com/claude-code)_
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
