Behavioral Changes in Apache Hive

Behavioral changes denote a marked change in behavior from the previously released version to this version of Apache Hive.

Cloudera ID:

CDPD-12108

Apache component

Hive

Summary:

Query results are returned unordered

Details:

Previous behavior:

When using the Hive Warehouse Connector in JDBC cluster mode, SELECT ... ORDER BY returned ordered results.

New behavior:

When using the Hive Warehouse Connector in JDBC cluster mode, SELECT ... ORDER BY returns unordered results. There is no guarantee on ordering when a SQL query is executed via "hive.sql()" . Since the order by action is inside the query sent to HiveServer, and therefore not known to Spark, the data in the relation is not ordered before being returned as a result.