reuvenlax commented on a change in pull request #15249:
URL: https://github.com/apache/beam/pull/15249#discussion_r684338356
##########
File path:
sdks/java/core/src/main/java/org/apache/beam/sdk/transforms/GroupIntoBatches.java
##########
@@ -398,11 +424,31 @@ public void processElement(
storedBatchSizeBytes.readLater();
}
- long num = storedBatchSize.read();
- if (maxBufferingDuration.isLongerThan(Duration.ZERO) && num == 1) {
- // This is the first element in batch. Start counting buffering time
if a limit was set.
- bufferingTimer.offset(maxBufferingDuration).setRelative();
+ long num;
+ if (maxBufferingDuration.isLongerThan(Duration.ZERO)) {
+ minBufferedTs.readLater();
+ num = storedBatchSize.read();
+
+ long oldOutputTs =
+ MoreObjects.firstNonNull(
+ minBufferedTs.read(),
BoundedWindow.TIMESTAMP_MAX_VALUE.getMillis());
+ minBufferedTs.add(elementTs.getMillis());
+ // If this is the first element in the batch or if the timer's output
timestamp needs
+ // modifying, then set a
+ // timer.
+ if (num == 1 || minBufferedTs.read() != oldOutputTs) {
+ long targetTs =
+ MoreObjects.firstNonNull(
+ timerTs.read(),
+ bufferingTimer.getCurrentRelativeTime().getMillis()
Review comment:
That is what's happening - getCurrentRelativeTime is basically current
time. We don't want to query the OS for current time because then unit tests
won't work. Unit tests (that use TestStream) inject a fake clock so that they
can control the advancement of time, and that won't work if we query the OS
itself.
--
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]
