jihoonson commented on a change in pull request #10171:
URL: https://github.com/apache/druid/pull/10171#discussion_r454735103
##########
File path:
processing/src/main/java/org/apache/druid/query/timeseries/TimeseriesQueryQueryToolChest.java
##########
@@ -138,7 +139,9 @@ public
TimeseriesQueryQueryToolChest(TimeseriesQueryMetricsFactory queryMetricsF
final Sequence<Result<TimeseriesResultValue>> finalSequence;
- if (query.getGranularity().equals(Granularities.ALL) &&
!query.isSkipEmptyBuckets()) {
+ if (query.getGranularity().equals(Granularities.ALL) &&
Review comment:
To be more sql-compliant, the timeseries query returns something if
there is no grouping key even when there is nothing to return. However, this
behavior is not preferred I think when `bySegment` is set because, `bySegment`
is used for caching in historicals or debugging where exact results are
supposed to be returned. When caching is enabled, it will populate empty result
in historicals after this change, but the broker will return the correct result
since `bySegment` won't be set for it. I added some comment for it.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
