Oracle Joins SQL-On-Big-Data Bandwagon

Oracle announced Oracle Big Data SQL on Tuesday, a new tool designed to run SQL queries across Hadoop, NoSQL, and Oracle Database, minimizing data movement while breaking down the silos of otherwise separate platforms.

Oracle currently offers a number of connectors for moving data between platforms including Oracle Database, Hadoop, and the Oracle NoSQL database. Oracle Big Data SQL goes the next step of supporting SQL-based querying across platforms. The new feature runs on the Oracle Big Data Appliance, an engineered system that includes Oracle Sun hardware, Cloudera's Hadoop Software distribution, and the Oracle NoSQL Database.

[Want more on multi-faceted big-data analysis? Read Will Spark, Google Dataflow Steal Hadoop's Thunder?]

Multiple vendors have introduced unified access layers and SQL-on-Hadoop options. The idea is to eliminate time- and labor-intensive data movement while taking advantage of the skills of workers who are familiar with the SQL language.

Teradata's Unified Data Architecture and Query Grid, for example, supports SQL-based querying across Hadoop, MongoDB, and, soon, the Teradata Aster data-discovery platform, which supports SQL-based MapReduce, time-series, graph, and R-based analysis. Microsoft offers PolyBase, a feature of Microsoft SQL Server Parallel Data Warehouse, that supports combined analysis of relational and non-relational data. And then there are database vendors, including Actian, Exasol, InfiniDB, and Pivotal, that have ported existing (SQL) relational databases to run on top of Hadoop.

Based on Oracle's descriptions, Oracle Big Data SQL sounds similar to Microsoft's PolyBase or Teradata's Query Grid features. Additional details are to be shared in a one-hour Webcast to be presented by Oracle Executive VP Andrew Mendelsohn on Tuesday. Oracle already offers a SQL-on-Hadoop option by way of Cloudera Impala, which is included with the Cloudera software shipped with the Oracle Big Data appliance.

"Impala enables customers to query data with SQL natively and efficiently in Hadoop," said Cloudera founder, chief strategy officer, and chairman Mike Olsen in a statement from Oracle. "For customers who need to query and analyze data residing in both Hadoop and Oracle Database, Oracle Big Data SQL offers support for HDFS, preserving existing SQL skills and security policies, and making it easier to integrate Hadoop with existing Oracle infrastructure.”

The big question with Oracle Big Data SQL is just what type of data analysis it will support other than conventional SQL? MapReduce was initially the key form of data processing against unstructured data that set the Hadoop platform apart. With Hadoop 2.0, the big data community is looking beyond the complexity and batch-oriented nature of MapReduce. Alternatives like Spark support machine learning, graph processing, and streaming analysis as well as SQL. With Aster, Teradata uses SQL to support MapReduce, graph, and R-based analyses as well as SQL.

We'll be watching Tuesday's presentation from Oracle to report on just what Oracle Big Data SQL can do.

InformationWeek's June Must Reads is a compendium of our best recent coverage of big data. Find out one CIO's take on what's driving big data, key points on platform considerations, why a recent White House report on the topic has earned praise and skepticism, and much more.