cshuo commented on code in PR #13592:
URL: https://github.com/apache/hudi/pull/13592#discussion_r2224545830
##########
hudi-flink-datasource/hudi-flink/src/main/java/org/apache/hudi/source/StreamReadMonitoringFunction.java:
##########
@@ -213,28 +244,47 @@ public void
monitorDirAndForwardSplits(SourceContext<MergeOnReadInputSplit> cont
// table does not exist
return;
}
- IncrementalInputSplits.Result result =
- incrementalInputSplits.inputSplits(metaClient, this.issuedOffset,
this.cdcEnabled);
+
+ IncrementalInputSplits.Result result = remainingSplits.isEmpty()
+ ? incrementalInputSplits.inputSplits(metaClient, this.issuedOffset,
this.cdcEnabled)
Review Comment:
IIUC, it's here where performance bottleneck lies, so the problem still
exists w/ this PR if one commit contains excessive amount of input splits?
##########
hudi-flink-datasource/hudi-flink/src/main/java/org/apache/hudi/source/StreamReadMonitoringFunction.java:
##########
@@ -97,8 +102,14 @@ public class StreamReadMonitoringFunction
private String issuedOffset;
+ private int totalSplits = -1;
+
+ private List<MergeOnReadInputSplit> remainingSplits = new ArrayList<>();
Review Comment:
transient
--
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]
