11.102. Release 0.139

Dynamic Split Concurrency

The number of running leaf splits per query is now dynamically adjusted to improveoverall cluster throughput. task.initial-splits-per-node can be used to setthe initial number of splits, and task.split-concurrency-adjustment-intervalcan be used to change how frequently adjustments happen. The session propertiesinitial_splits_per_node and split_concurrency_adjustment_interval canalso be used.

General Changes

  • Fix planning bug that causes some joins to not be redistributed whendistributed-joins-enabled is true.
  • Fix rare leak of stage objects and tasks for queries using LIMIT.
  • Add experimental task.join-concurrency config which can be used to increaseconcurrency for the probe side of joins.

Hive Changes

  • Remove cursor-based readers for ORC and DWRF file formats, as they have beenreplaced by page-based readers.
  • Fix creating tables on S3 with CREATE TABLE AS.