Writing Data to Hive
Core Storm and Trident APIs support streaming data directly to Apache Hive using Hive transactions. Data committed in a transaction is immediately available to Hive queries from other Hive clients. You can stream data to existing table partitions, or configure the streaming Hive bolt to dynamically create desired table partitions.
Use the following steps to perform this procedure:
Instantiate an implementation of the
HiveMapperInterface.Instantiate a
HiveOptionsclass with theHiveMapperimplementation.Instantiate a
HiveBoltwith theHiveOptionsclass.
![[Note]](../common/images/admon/note.png) | Note |
|---|---|
Currently, data may be streamed only into bucketed tables using the ORC file format. |