jingz-db commented on code in PR #47528:
URL: https://github.com/apache/spark/pull/47528#discussion_r1699101433
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/OperatorStateMetadata.scala:
##########
@@ -166,18 +169,28 @@ object OperatorStateMetadataUtils extends Logging {
s"version=${operatorStateMetadata.version}")
}
}
+
+ def getLastOffsetBatch(session: SparkSession, checkpointLocation: String):
Long = {
Review Comment:
Just thinking out loud here, so we are reading metadata from latest batch in
the offsetlog. Suppose a query failed before committing the batch and write
nothing to the metadata path, and then we start a metadata reader from it:
could there be situation where a batchId exists in offsetLog but there is
actually no metadata written at this batchId?
--
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]
