Github user JamesRTaylor commented on a diff in the pull request:
https://github.com/apache/phoenix/pull/308#discussion_r203879864
--- Diff:
phoenix-core/src/main/java/org/apache/phoenix/execute/ClientAggregatePlan.java
---
@@ -135,17 +142,24 @@ public ResultIterator iterator(ParallelScanGrouper
scanGrouper, Scan scan) throw
aggResultIterator = new
ClientUngroupedAggregatingResultIterator(LookAheadResultIterator.wrap(iterator),
serverAggregators);
aggResultIterator = new
UngroupedAggregatingResultIterator(LookAheadResultIterator.wrap(aggResultIterator),
clientAggregators);
} else {
- if (!groupBy.isOrderPreserving()) {
- int thresholdBytes =
context.getConnection().getQueryServices().getProps().getInt(
- QueryServices.SPOOL_THRESHOLD_BYTES_ATTRIB,
QueryServicesOptions.DEFAULT_SPOOL_THRESHOLD_BYTES);
- List<Expression> keyExpressions =
groupBy.getKeyExpressions();
+ List<Expression> keyExpressions = groupBy.getKeyExpressions();
+ if (groupBy.isOrderPreserving()) {
+ aggResultIterator = new
ClientGroupedAggregatingResultIterator(LookAheadResultIterator.wrap(iterator),
serverAggregators, keyExpressions);
--- End diff --
Pass through context here too to ClientGroupedAggregatingResultIterator as
you'll need it to get the memory manager.
---
