1996fanrui commented on code in PR #19723:
URL: https://github.com/apache/flink/pull/19723#discussion_r880013345


##########
flink-runtime/src/main/java/org/apache/flink/runtime/io/network/partition/PipelinedSubpartition.java:
##########
@@ -240,9 +260,158 @@ private boolean processPriorityBuffer(BufferConsumer 
bufferConsumer, int partial
                         inflightBuffers.toArray(new Buffer[0]));
             }
         }
-        return numPriorityElements == 1
-                && !isBlocked; // if subpartition is blocked then downstream 
doesn't expect any
-        // notifications
+        return needNotifyPriorityEvent();
+    }
+
+    // It just be called after add priorityEvent.
+    private boolean needNotifyPriorityEvent() {
+        assert Thread.holdsLock(buffers);
+        // if subpartition is blocked then downstream doesn't expect any 
notifications
+        return buffers.getNumPriorityElements() == 1 && !isBlocked;
+    }
+
+    private void processTimeoutableCheckpointBarrier(BufferConsumer 
bufferConsumer) {
+        CheckpointBarrier barrier = 
parseAndCheckTimeoutableCheckpointBarrier(bufferConsumer);
+        channelStateWriter.addOutputDataFuture(
+                barrier.getId(),
+                subpartitionInfo,
+                ChannelStateWriter.SEQUENCE_NUMBER_UNKNOWN,
+                createChannelStateFuture(barrier.getId()));
+    }
+
+    private void completeTimeoutableCheckpointBarrier(BufferConsumer 
bufferConsumer) {
+        CheckpointBarrier barrier = 
parseAndCheckTimeoutableCheckpointBarrier(bufferConsumer);
+        if (channelStateFutureIsAvailable(barrier.getId())) {
+            completeChannelStateFuture(Collections.emptyList(), null);
+        }
+    }

Review Comment:
   Actually, it may also be false here.
   
   You mentioned in a [previous 
comment](https://github.com/apache/flink/pull/19723#discussion_r875800201). It 
happens when a checkpoint that has been aborted previously.
   
   BTW, for these 2 comments, I'll add a brief description to the code 
explaining why we can't checkState, to prevent it from being corrected later.



-- 
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: issues-unsubscr...@flink.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org

Reply via email to