Pentaho acquisition by Hitachi Data Systems is now old story. The product, in its Community version, is still not influenced by the new roadmap, while the Enterprise version is actually steering towards @scale solutions. Two articles ought to be read:
- Streamlined Data Refinery aka a continuous process aimed at analyzing data, starting from the very input data without the need for an intermediate DB/DataWarehouse in charge of IT. It really looks like old self-service BI, actually the difference lies in the Hadoop common layer, which is now sort of common layer/transport of information among all suite products. The development and life cycle is still web based, thus reducing the chaos caused by objects living in end user desktops.
- The Hitachi appliance for Pentaho BA Server so the corresponding infrastructure for running the above.
What on the IoT side? Till now just rumors from developers of Kettle/Pentaho Data Integration (aka Matt Casters and his hangouts with the community) about efforts for enabling Metadata Injection on more and more steps… but this is just a starting point to let trasformations get data from mutable inputs. It’s like if the ETL were moving to a noSql record structure…
Summing thigs up: we’re still with a regular Big Data Analytics solution, but, since it’s now production ready, it’s easy to guess next step: IoT devices directly injecting messages on this platform. the new Hitachi standard de facto.