A lot of folks tend to think of the management of Hadoop from an IT perspective as a disruptive technology that will wind up replacing all manner of data management systems. In reality, emerging technologies such as Hadoop generally get deployed alongside existing IT investments.
As a cost-effective platform for managing massive amounts of unstructured data, Tasso Argyros, co-president of Teradata Aster, says Hadoop provides IT organizations with a platform that allows them to pre-process a huge amount of data. But Hadoop is based on a micro batch-processing architecture that doesn't scale all that well from a performance perspective or seamlessly support existing SQL-based applications.
To address that specific issue, Teradata today announced that a new Aster SQL-H offering now offers business analysts a bridge between standard business intelligence (BI) applications and multi-structured Big Data stored in the Hadoop Distributed File Systems (HDFS).
Argyros says end users are looking for ways to extend the value of their existing investments in SQL-based analytics applications. By allowing them to query Hadoop data, Aster SQL-H also makes it easier to access the Aster MapReduce platform, which includes over 50 pre-built MapReduce analytical applications and a SQL-MapReduce interface.
The issue that Teredata is trying to tackle is the cumbersome nature of a MapReduce interface that tracks its lineage back to the LISP programming language. While there is a movement afoot to build applications directly on Hadoop, the fact remains that the mass of data management investments in the enterprise today are based on SQL. For that reason, Teradata is working with partners such as Hortonworks to integrate its Aster Data massively parallel database appliance with software that makes it easier to manage Hadoop deployments, says Argyros.
Ultimately, Argyros says Hadoop will serve to increase the value of existing investments in data management technologies by combining a low-cost approach to storing massive amounts of data with platforms such as the Aster Data database appliance that is capable of processing that data at scale. The challenge facing IT organizations is to figure out how to manage a rapidly expanding ecosystem of data management technologies that are going to be collectively needed to manage Big Data in all its forms.