Currently, hadoop-connectors are implemented by using new MapReduce API but some applications like Hive only supports Input/OutputFormat with the old MapRed interface.
So far I don't see there is a requirement to support Hive, but just create this issue to track this potential effort. It's a low priority now.
Currently, hadoop-connectors are implemented by using new MapReduce API but some applications like Hive only supports Input/OutputFormat with the old MapRed interface.
So far I don't see there is a requirement to support Hive, but just create this issue to track this potential effort. It's a low priority now.