Automatically ingest new data from hadoop to druid?

What should be the config if my data in hdfs is partitioned by yyyy/MM/dd/hh format. Does druid ingestion task automatically load new data periodically when available?