T
Tim Kirwan
Guest
- What is being requested: Method to reduce the number of MapReduce jobs created when performing INSERT of multiple rows via parameterized insert - Reason for request: Opened case 00328667 when customer observed that a parameterized insert created a MapReduce job for each row inserted. Results of case were: The behavior you are seeing is expected behavior when using parameter arrays. The driver performs individual inserts because the Hive protocol does not have a mechanism to support parameter arrays. Since the parameter array has 26 rows, the driver performs 26 inserts, which results in 26 mapreduce jobs. This same behavior can be seen using Beeline. - Possible workarounds: No alternatives or workarounds exist for this requirement.
Continue reading...
Continue reading...