Follow Datanami:
October 21, 2013

Teradata Moving to the Cloud

Alex Woodie

Teradata is under the gun to grow its business and compete more effectively with big data startups after reporting disappointing financial results earlier this month. To that end, the company today made several announcements at its annual user conference, which is taking place this week in Dallas, Texas, including putting its Teradata and Aster offerings available in the cloud, and support for JSON in its database.

The Teradata Cloud will eventually give customers access to the full spectrum of the company’s technologies, including its core relational database management system (RDBMs) and its Aster data discovery software, along with Hadoop, ETL, and surrounding management tools.

The first offering under the new Teradata Cloud banner is Data Warehouse as a Service, which is available today in the United States. This offering gives customers access to a hosted copy of the Teradata RDBMS, along with ETL and business intelligence (BI) tools. It will also include “industry starter kits” aimed at helping customers get started with data analytic processes, such as data modeling, data loading, and BI reporting. In addition to providing the software, Teradata takes care of all the hardware and management tasks, and bills customers on a monthly basis.

The BI giant plans to add more offerings to its new cloud business. In early 2014, the company plans to debut Discovery as a Service and Data Management as a Service. The Discovery as a Service offering will be built on the Aster discovery platform, and will enable customers to take advantage of Aster’s SQL, MapReduce, and graph processing engines from the cloud. The Data Management as a Service offering will deliver Hadoop processing from a cloud.

Shawn Rogers, a vice president with the analytic consulting company Enterprise Management Associates, says the market for cloud analytics will grow over the next three to five years. “With this new cloud offer, Teradata delivers a flexible, scalable option for a complete set of cloud-based analytics, industry expertise, and consulting services,” he says.

One of the early adopters of the Teradata cloud is Netflix, the big data darling that everybody wants on their customer list. With reams of customer preference data being generated every second, it’s no wonder that Netflix has its hands in a lot of big data analytic cookie jars. While Netflix is widely known to use the DataStax distribution of the Cassandra database as its production recommendation engine,  the company is actively perusing the biggest, juiciest morsels to give it an analytical edge.

“We will use Teradata Cloud as our database platform for high concurrency and mixed workloads, which require sophisticated integration and processing at extremely high speeds,” Netflix director of data platforms Kurt Brown stated. “With hundreds of analysts examining billions of data points we felt that Teradata best met our needs for workload management, scalability and concurrent users.”

In other news, the company announced that it will support Java Script Object Notation (JSON), expected to occur with the next major release of the Teradata Database in 2Q14.

Much of the modern Web, and the Internet as a whole, is being underpinned by JSON, which is less structured than other data types but more flexible too. Traditional relational database vendors, including Oracle, IBM, and Teradata with its eponymous database, have been scrambling to add JSON support to catch up with NoSQL database vendors, many of which already support JSON.

Teradata says the addition of JSON to the warehouse will allow customers to use their Teradata data warehouses as analytic hubs to “gain business value from the Internet of Things.” “The Internet of Things will drive the next wave of data growth and it has the potential to change the world, just as the Internet did,” Teradata Labs president Scott Gnau said in a statement.

Teradata plans to store the entire JSON object directly in a column, which the company says will enable business users to access them more easily. Also, Teradata plans to take advantage of JSON’s capability to store multiple documents inside one big document, which will enable the IT company to store a complete history of information.

For example, when storing information for a location-based ad warehouse, JSON will enable Teradata customers to store a complete history of a consumer’s locations, responses to geo-based ads, and activity before and after seeing the ads, Teradata says.

The company made several other announcements at the show, including:

  • its new Extreme Data Platform 1700, a new offering that delivers SQL-based processing of data, for $2,000 per terabyte of compressed data.
  • an expanded partnership with SAS and Hortonworks to sell something called the Analytics Advantage Program with Hadoop
  • new offerings for marketers, including Customer Interaction Manger in the Cloud and Integrated Marketing Management (IMM) in the cloud.

Related Items:

Teradata Adds Graph Engine to the Data Discovery Mix with Aster 6

Teradata Gets In Deep with R

Teradata Delivers Industry’s First Flexible, Comprehensive Hadoop Portfolio