heyihong commented on code in PR #53801:
URL: https://github.com/apache/spark/pull/53801#discussion_r2718867558
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/SQLExecution.scala:
##########
@@ -239,6 +239,8 @@ object SQLExecution extends Logging {
event.qe = queryExecution
event.executionFailure = ex
sc.listenerBus.post(event)
+
+ sparkSession.observationManager.tryComplete(queryExecution)
Review Comment:
Thanks for explaining. I think for connect mode, there is no difference
because Observation is waiting for the results in the same thread. For classic
mode, there are trade-offs since it may also block other async tasks. It may
make sense to introduce a way to do all these things in parallel instead of
sequentially if latency is important.
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
