Page History

...

When true, this turns on dynamic partition pruning for the Spark engine, so that joins on partition keys will be processed by writing to a temporary HDFS file, and read later for removing unnecessary partitions.

hive.spark.dynamic.partition.pruning.map.join.only

• Default Value: false
• Added In: Hive 3.0.0 with HIVE-16998

Similar to hive.spark.dynamic.partition.pruning, but only enables DPP if the join on the partitioned table can be converted to a map-join.

hive.spark.dynamic.partition.pruning.max.data.size

• Default Value: 100MB
• Added In: Hive 1.3.0 with HIVE-9152

...