Created 03-13-2017 01:18 AM
Documentation for PutHiveStreaming requires flow file to be in avro format. My understanding is hive streaming only support ORC format. When puthivestreaming is used, does it convert avro to orc prior to inserting into hive table? trying to understand the functionality.
Created 03-13-2017 02:27 AM
It does convert Avro to ORC, but ironically it does so by converting the Avro to JSON first. We should allow ORC (if prudent, possibly after a ConvertAvroToORC processor) and JSON as valid input formats (perhaps via the mime.type attribute?) to PutHiveStreaming. Do you mind writing up a Jira for this? Please and thank you 🙂
Created 03-13-2017 02:27 AM
It does convert Avro to ORC, but ironically it does so by converting the Avro to JSON first. We should allow ORC (if prudent, possibly after a ConvertAvroToORC processor) and JSON as valid input formats (perhaps via the mime.type attribute?) to PutHiveStreaming. Do you mind writing up a Jira for this? Please and thank you 🙂
Created 03-20-2017 04:39 AM
Please see https://issues.apache.org/jira/browse/NIFI-3625.