I have used Hortonworks HDP in my R&D phase and found it quite easy to use. Currently I am exploring a Banking Use case and I am facing a tough decision regarding choice of technology. Most of Bank’s data is in Oracle RDBMS. Oracle has a product ‘Oracle R Enterprise’ which allows distributed processing of structured data using R. In this case, why would anyone choose to have the structured data loaded first into the HDFS and then process it through MapReduce (Hadoop)?. I have seen HortonWork’s reference architecture for Finance and Banking, in which data from data warehouses and other databases are Ingested into HDP. Can someone please help me understand how do we choose which technology is better in this case?
↧