dawidwys commented on a change in pull request #15055:
URL: https://github.com/apache/flink/pull/15055#discussion_r662811609
##########
File path:
flink-runtime/src/main/java/org/apache/flink/runtime/io/network/partition/consumer/RecoveredInputChannel.java
##########
@@ -205,6 +205,12 @@ public void resumeConsumption() {
throw new UnsupportedOperationException("RecoveredInputChannel should
never be blocked.");
}
+ @Override
+ public void acknowledgeAllRecordsProcessed() throws IOException {
+ throw new UnsupportedOperationException(
+ "RecoveredInputChannel should not need acknowledge all records
processed.");
+ }
+
Review comment:
Ok, so what I said in
https://github.com/apache/flink/pull/15055#discussion_r662031701 is actually
already the current state. I second @gaoyunhaii it should work correctly the
way it is. Basically, the `EndOfUserRecordsEvent` will never be in the
input/output buffers during restoring.
As for an ITCase, I think it makes sense once we have all building blocks in
place.
--
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]
