pnowojski commented on code in PR #28689:
URL: https://github.com/apache/flink/pull/28689#discussion_r3577045341
##########
flink-runtime/src/test/java/org/apache/flink/streaming/api/operators/SourceOperatorSplitWatermarkAlignmentTest.java:
##########
@@ -569,6 +569,58 @@ void testAlignmentCheckIsDeferredForIdleSplits() throws
Exception {
0L,
operator.getSplitMetricGroup(split0.splitId()).getAccumulatedPausedTime());
}
+ @Test
+ void testPausedIdleSplitsCanBeResumedByAlignmentCheck() throws Exception {
+ final long idleTimeout = 100;
+ final MockSourceReader sourceReader =
+ new MockSourceReader(WaitingForSplits.DO_NOT_WAIT_FOR_SPLITS,
true, true);
+ final TestProcessingTimeService processingTimeService = new
TestProcessingTimeService();
+ final SourceOperator<Integer, MockSourceSplit> operator =
+ createAndOpenSourceOperatorWithIdleness(
+ sourceReader, processingTimeService, idleTimeout);
+
+ final MockSourceSplit split0 = new MockSourceSplit(0, 0, 10);
+ final int allowedWatermark4 = 4;
+ final int allowedWatermark7 = 7;
+ split0.addRecord(4);
+ split0.addRecord(5);
+ split0.addRecord(6);
+ split0.addRecord(7);
+ split0.addRecord(8);
+ operator.handleOperatorEvent(
+ new AddSplitEvent<>(Arrays.asList(split0), new
MockSourceSplitSerializer()));
+ final CollectingDataOutput<Integer> actualOutput = new
CollectingDataOutput<>();
+
+ // Emit enough records to fill the sampler buffer
+ for (int i = 0; i < WATERMARK_ALIGNMENT_BUFFER_SIZE.defaultValue();
i++) {
+ operator.emitNext(actualOutput);
+ processingTimeService.advance(idleTimeout - 1);
+ }
+ sampleAllWatermarks(processingTimeService);
+ assertOutput(actualOutput, Arrays.asList(4, 5, 6));
+
+ // Alignment check fires and pauses the split
+ operator.handleOperatorEvent(new
WatermarkAlignmentEvent(allowedWatermark4));
+
assertThat(operator.getSplitMetricGroup(split0.splitId()).isPaused()).isTrue();
+ assertThat(sourceReader.getPausedSplits()).containsExactly("0");
+ assertOutput(actualOutput, Arrays.asList(4, 5, 6));
+
+ // Normally idlenessTimer can't elapse while the split is paused
+ // So calling it manually to simulate a race condition
Review Comment:
I'm asking, because I wonder if we can replicate that bug without calling
explicitly `operator.updateCurrentSplitIdle`, by just using api as it was
intended to be used?
So for example, can not we just flip the order, first advance manual clock
so that idle timeout is triggered, then send `WatermarkAlignmentEvent` to the
operator to trigger `pauseOrResumeSplits`?
--
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]
