Is raw data still needed after segment creation?


I batch ingested json files. I configured my deep storage to be in /data/deep and the raw data is read from /data/raw (firehorse.baseDir parameter in my indexing task).

My question is now, if this raw file from the baseDir is still needed after the segment is created. Can I delete it or will this cause any problems to Druid.

Thanks for your help


The raw file is not needed by Druid anymore.

That said, we generally recommend that a production system maintain
raw data in some warehouse somewhere so that it can be re-indexed,
adjusted, and just generally processed in ways that Druid might not