Github user JoshRosen commented on a diff in the pull request: https://github.com/apache/spark/pull/9067#discussion_r42060953 --- Diff: sql/core/src/main/scala/org/apache/spark/sql/execution/aggregate/TungstenAggregationIterator.scala --- @@ -470,12 +484,27 @@ class TungstenAggregationIterator( // Process the rest of input rows. while (inputIter.hasNext) { - val newInput = inputIter.next() - numInputRows += 1 - val groupingKey = groupProjection.apply(newInput) - buffer.copyFrom(initialAggregationBuffer) - processRow(buffer, newInput) - externalSorter.insertKV(groupingKey, buffer) + var newHashMap = initHashMap() --- End diff -- I think you might actually want to re-assign to the original `hashMap` by making it into a `var`. This will let us ensure that `UnsafeAggregationIterator.free()` is able to free all memory after failures.
--- 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. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org