Analyst's Take: Teradata 12.0 Counters the Appliances

More efficient processing, workload management and master data management help handle mixed workloads.
To speed up queries and updates, the 12.0 release offers a feature called multilevel partition primary index, which lets users place indexes on particular columns to see and compare data — for example, from specific time periods and sales regions — without forcing a time-consuming and process-intensive scan of entire tables that could contain terabytes of data. The new release also improves the system's ability to collect statistics about query plans and use the database engine's optimizer to rewrite queries for better performance. It also provides event-based workload management, which Teradata describes as one of the most important elements of the new release. Workload management automates steps that database administrators otherwise have to perform manually to make sure that resources are available for important users and queries and are released for other workloads when no longer needed.

Teradata is elevating the profile of master data management as well. The company is developing its MDM software on technology it gained and controls through a 2005 partnership with i2 Technologies. Teradata 12.0 manages MDM centrally from the enterprise data warehouse and employs the technology much like an extraction, transformation and loading (ETL) process to ensure data quality and consistency of definitions. The company is expanding its implementation of MDM to include version management so organizations can see what their master data looked like at specific points in time. Teradata also is integrating MDM with data modeling tools and publishing services so that new and existing definitions can be consolidated, managed and communicated more effectively.

Market Impact

Competition is intensifying in the data warehouse market as organizations evaluate technology to support both operational BI and real-time analytics. Data warehouse appliances from DATAllegro, Greenplum and Netezza, as well as configuration packages from Hewlett-Packard, IBM, Oracle and others, are diversifying the options. Thus, to protect and expand its customer base, Teradata needs to show how it can enable enterprise data warehouses to step up to the mixed workload challenge. The company also needs to demonstrate information management and workload management capabilities that enable customers to manage increasing demands efficiently and cost-effectively. The 12.0 release appears to satisfy these needs. Nonetheless, if the rollout of version 12.0 is to proceed successfully Teradata will have to present current and prospective customers with a strong case for deploying its technology to handle both operational BI and real-time analytics.


Ventana Research believes that companies seeking to expand traditional data warehousing to enable operational BI and real-time analytics should evaluate Teradata Enterprise Data Warehouse 12.0. Current customers that are considering whether to deploy data warehouse appliances rather than increase the number of their Teradata systems should look carefully at 12.0's data management features and processing improvements to determine whether it enables them to increase the performance of queries for specific reports, user groups or business processes cost-effectively without adding new nodes to the system.

Organizations today are fortunate that heightened market competition is putting pressure not only on Teradata but on other established database vendors to make significant improvements to their technology rather than offer only incremental maintenance changes. Whether they are upgrading to new releases or deploying appliances, however, it is important that organizations make decisions regarding data warehousing technology within the context of a comprehensive information management strategy.

David Stodder is vice president and research director - information management and IT performance management at Ventana Research. He was previously editor-in-chief and editorial director of Intelligent Enterprise. Write him at [email protected]