shunping commented on code in PR #39487:
URL: https://github.com/apache/beam/pull/39487#discussion_r3676393013
##########
runners/google-cloud-dataflow-java/worker/src/main/java/org/apache/beam/runners/dataflow/worker/DataflowOutputCounter.java:
##########
@@ -63,11 +64,27 @@ public void update(Object elem) throws Exception {
objectAndByteCounter.update(elem);
long windowsSize = ((WindowedValue<?>) elem).getWindows().size();
if (windowsSize == 0) {
- // GroupingShuffleReader produces ValueInEmptyWindows.
- // For now, we count the element at least once to keep the current
counter
- // behavior.
- elementCount.addValue(1L);
+ // ValueInEmptyWindows occurs when processing shuffle/streaming work
items
+ Object value = ((WindowedValue<?>) elem).getValue();
+ if (value instanceof KeyedWorkItem<?, ?>) {
+ // KeyedWorkItem wrapped in ValueInEmptyWindows
+ // (e.g. WindowingWindmillReader for Streaming GBK)
+ KeyedWorkItem<?, ?> keyedWorkItem = (KeyedWorkItem<?, ?>) value;
+ long totalElementCount = 0;
+ // Iterate only through elementsIterable and ignore timers in
KeyedWorkItem.
+ for (WindowedValue<?> element : keyedWorkItem.elementsIterable()) {
Review Comment:
Done. Added `elementWindowsIterable` to decode window metadata only. PTAL
--
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]
