neilramaswamy commented on code in PR #47475:
URL: https://github.com/apache/spark/pull/47475#discussion_r1690693797
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/StateStore.scala:
##########
@@ -881,7 +862,8 @@ object StateStore extends Logging {
case NonFatal(e) =>
logWarning(log"Error managing
${MDC(LogKeys.STATE_STORE_PROVIDER, provider)}, " +
log"stopping management thread", e)
- threadPoolException.set(e)
Review Comment:
Nope, there's no mechanism for that. But pulling on that thread, let's say
that the maint task does fall behind and can never make progress. Then, the
user will see that their checkpoint location never gets cleaned up. And, if
they fail, due to not having a snapshot (in the case of changelog
checkpointing), they will have slower recovery times. Is that correct?
--
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]
