This message was deleted.

I am not familiar with the Hadoop based jobs, so I can't help directly there. But have you tried using SQL Based ingestion to do this directly in Druid?

Hi Sergio, I'm currently running Druid native batch jobs to do the re-indexing but I wanted to move/test re-indexing with Hadoop to speed up re-indexing. I don't want to replace the content of the existent datasource with new data. In other words, I just want to re-index data from realtime datasource to fill the historical one with hadoop (which is suppose to be faster than Druid native batch ingestion jobs).