New York-based marketing analytics company Razorfish plans to pilot the new Teradata-Aster Big Analytics Appliance, saying the integration of the Teradata-Aster database with MapReduce functions and the Apache Hadoop distributed file system will make the company more efficient and enable it to streamline how it manages the technologies.
Unveiled last month, the Teradata-Aster Big Analytics Appliance integrates Teradata’s Aster database, the analytics processing of MapReduce and the open source data loading of Hadoop via Teradata’s partnership with Hortonworks.
“Right now at Razorfish, the management of Teradata-Aster, MapReduce and Hadoop are done separately,” said Matthew Comstock, Razorfish’s vice president of business intelligence. “The appliance offers a way for us to simplify this environment and gain more control. It should really reduce a lot of the manual batch processing we’ve had to do in the past.”
Comstock said that Razorfish has seen faster analytics performance by running Teradata-Aster, MapReduce and Hadoop, but managing all those environments was very time consuming.
One major advantage the appliance offers is that it comes packed with Viewpoint, Teradata-Aster’s web-based management tool. But central to the new appliance is technology called SQL-H, software that offers business analysts a bridge between standard business intelligence applications and access to big datasets stored in Hadoop. Comstock said in the past, database architects would have to load MapReduce functions into Apache Hadoop and take that output and load it into Teradata-Aster.
“SQL-H automates all that manual work,” Comstock said. “With the appliance, we can have all the functionality of Teradata-Aster, MapReduce and Hadoop in one system.”
In introducing the appliance, Tasso Argyros, copresident of Teradata-Aster, said the company was looking to meet a need in the marketplace to make big data analytics easier to manage. The appliance “will help our customers overcome the perception that big data analytics technology is hard to deploy and achieve business value,” he said.
IDC Analyst Dan Vesset, who attended the Teradata Partners Conference, said the new appliance addresses shortcomings that have plagued the industry to date, most notably reducing the need to have Hadoop specialists on staff.
“There’s certainly a place for this kind of device,” he said. “Many internal organizations simply don’t have the skills to set up a Hadoop cluster and this device does that for them. That, in and of itself, has value. The appliance allows IT departments to collocate Aster with Hadoop, and if you have the SQL skills you can do analysis based on MapReduce.”
Vesset said the appliance may be especially useful to existing Teradata and Aster customers who are new to Hadoop. Another of its real advantages, he said, is that business analysts can now integrate and analyze the structured data stored in the Teradata relational database and unstructured and semi-structured data contained in web logs and social media traffic.
Chris Twogood, Teradata’s vice president of product and services marketing, said the new appliance was designed to work with existing business intelligence applications over the new equipment. “You don’t have to go out and buy new business intelligence or data visualization tools,” said Twogood.
Razorfish’s Comstock confimed Twogood’s claim about companies using existing tools, saying that Razorfish still runs Tableau for data visualization.
The Teradata-Aster appliance can be configured to store a maximum of 5 petabytes of uncompressed user data for Aster and up to 10 petabytes of uncompressed user data for Hadoop. It also comes with more than 50 standard MapReduce analytics functions. Teradata hopes to expand the number of apps through internal development; but stealing a page from the open source world, wants the user community to contribute new MapReduce apps as well.