Yes for us Spark is being used for just ETL. It's internal data though. It's at the scale where no single machine could ever handle it though. It's replacing

Yes for us Spark is being used for just ETL. It's internal data though. It's at the scale where no single machine could ever handle it though. It's replacing large proprietary systems like mainframes and teradata. We wrote a framework to heavily simplify writing these ETL flows. No developer of the ETL flows is using Scala unless something custom is required. I wish we could opensource it.