Github user Ben-Zvi commented on a diff in the pull request:
https://github.com/apache/drill/pull/822#discussion_r118154367
--- Diff:
exec/java-exec/src/main/java/org/apache/drill/exec/record/RecordBatch.java ---
@@ -188,6 +188,18 @@
NOT_YET,
/**
+ * Restart reading incoming batches.
+ * <p>
+ * Prior calls to {@link #next()} returned data after all incoming
batches were processed.
+ * But with this status, need to restart reading incoming
RecordBatches.
+ * </p>
+ * <p>
+ * Currently only used internally by the Hash Aggregate.
+ * </p>
+ */
--- End diff --
The whole iterator driven next() model is very limiting. This change was
the simplest - re-invoking the next next() so that spilled partitions could be
processed from then on.
Trying to switch without returning from the next() first seems too complex.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
