11.39. Release 0.191

General Changes

  • Fix regression that could cause high CPU usage for join queries when dictionaryprocessing for joins is enabled.
  • Fix bit_count() for bits between 33 and 63.
  • The query.low-memory-killer.enabled config property has been replacedwith query.low-memory-killer.policy. Use total-reservation to continueusing the previous policy of killing the largest query. There is also a newpolicy, total-reservation-on-blocked-nodes, which kills the query thatis using the most memory on nodes that are out of memory (blocked).
  • Add support for grouped join execution. When both sides of a join have thesame table partitioning and the partitioning is addressable, partial datacan be loaded into memory at a time, making it possible to execute the joinwith less peak memory usage. The colocated join feature must be enabled withthe colocated-joins-enabled config property or the colocated_joinsession property, and the concurrent_lifespans_per_task session propertymust be specified.
  • Allow connectors to report the amount of physical written data.
  • Add ability to dynamically scale out the number of writer tasks ratherthan allocating a fixed number of tasks. Additional tasks are added when thethe average amount of physical data per writer is above a minimum threshold.Writer scaling can be enabled with the scale-writers config property orthe scale_writers session property. The minimum size can be set with thewriter-min-size config property or the writer_min_size session property.The tradeoff for writer scaling is that write queries can take longer to rundue to the decreased writer parallelism while the writer count ramps up.

Resource Groups Changes

  • Add query type to the exact match source selector in the DB resource group selectors.

CLI Changes

  • Improve display of values of the Geometry type.

Hive Changes

  • Add support for grouped join execution for Hive tables when bothsides of a join have the same bucketing property.
  • Report physical written data for the legacy RCFile writer, optimized RCFilewriter, and optimized ORC writer. These writers thus support writer scaling,which can both reduce the number of written files and create larger files.This is especially important for tables that have many small partitions, assmall files can take a disproportionately longer time to read.

Thrift Connector Changes

  • Add page size distribution metrics.

MySQL, PostgreSQL, Redshift, and SQL Server Changes

  • Fix querying information_schema.columns if there are tables withno columns or no supported columns.