Using phenix as Kylin's input source is doable (as we have a plugable arch which allow to substitute input source and output storage for cubing).
The question will be, will phenix suffice as Kylin's input source ? I mean today we use hive to flatten the star schema tables as well as to do data filtering/cleaning (for example apply a where condition to the fact table records), the hive job might easily engage thousands of mappers/reducers to accomplish the job. I'm not an expert on phenix so we just need to figure our if phenix provides similar capacities. -- Regards, *Bin Mahone | 马洪宾* Apache Kylin: http://kylin.io Github: https://github.com/binmahone
