You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@drill.apache.org by Matt <bs...@gmail.com> on 2015/03/05 23:17:51 UTC

Choosing storage for OLAP

The default Drill 0.7 install contains plugins for HBase and Hive. As 
someone not overly familiar with either, is there consensus or 
documentation on which is preferable for OLAP use with time-series data?

An example case would be data with a key, timestamp, and multiple 
numeric fact columns, where common ad-hoc queries select based on the 
key and a time range, and aggregate the fact columns.

The source data in this case are csv files, but it would seem that 
storage with indexes would be preferable.