Teradata Active Enterprise Data Warehouse

Every organization is different: Each has its own objectives and growth path. Consequently, companies have varying needs and uses for their data warehouse environments. Some use their data warehouse for a single, simple application, while others leverage an enterprise data warehouse (EDW) to analyze data in near real time to get the full value of Active Enterprise Intelligence capabilities.

The scope of data warehousing needs can be fulfilled and a compatible value level provided with the robust Teradata Purpose-Built Platform Family: Teradata Active Enterprise Data Warehouse, Teradata Data Warehouse Appliance, Teradata Data Mart Appliance and Teradata Extreme Data Appliance.

All in the family
Each member of the platform family comes ready to run and is purpose-built for a distinct data warehousing need, from basic reporting and analysis to full-fledged active data warehousing. (See figure 1, below.) The Teradata Database is the foundation for each family member, so applications can easily port from one platform to another without changing underlying structures, making the clients' knowledge and expertise platform-universal. The family tree comprises:

TERADATA ACTIVE ENTERPRISE DATA WAREHOUSE
This family head takes maximum advantage of the massively parallel processor (MPP) architecture, a long-standing key enabler for Teradata system capabilities. The EDW-based platform has four key attributes:
1. High-performance technology. Exploits industry-leading Quad Core Intel Processor technology in high-performance processing nodes and supports optimized and flexible configurations of enterprise-class disk drives
2. Scalability. Linearly increases capacity and performance from a one-node to a 1024-node system and from 2TB to more than 10 petabytes
3. Availability. Reduces or eliminates the impact of hardware failures with redundant and hot-swappable components. In an entire node failure, the system's unique clique architecture enables full performance continuity
4. Growth with investment protection. Supports multiple platform generations in a single system to facilitate expansion with the latest platform generations. This increases the return on investment (ROI) for the initial system as well as the organization's confidence in the entire system's extended life expectancy

By combining strategic and operational analysis with the advanced workload management feature, this mission-critical active data warehouse helps meet the demands of both back-office workers and front-line users.

TERADATA DATA WAREHOUSE APPLIANCE
The fixed-configuration solution integrates processing and storage in cabinet building blocks that arrive staged and tested, as with all family members. It is optimized for the high-performance analytics required for decision support and provides the basic workload management needed for this work. Teradata Data Warehouse Appliance is ideal for users who are new to data warehousing or who want to complement an existing data warehouse. For instance, the platform can be used as a cyclical data mart for testing applications before their implementation into the EDW. It can also be used to off-load the work associated with applications that are best deployed on a separate platform but are still dependent on the data within the EDW. The appliance features:
1. High-performance technology. Uses Quad Core Intel Xeon Processors in high-performance processing nodes and Teradata Database innovations for enhanced scan performance
2. Scalability. Meets the needs of entry-level scalability, from a two-node to a 44-node system, (one to 11 integrated cabinets) with compressed user data capacities ranging from 6.3TB to 140TB (based on a typical compression factor of 40%)
3. Availability. Provides high availability with the Teradata node failover clique architecture, RAID-1 disk storage, and redundant and hot-swappable hardware

TERADATA DATA MART APPLIANCE
This cost-effective vehicle delivers the power and capabilities of a Teradata data warehouse to small enterprises. It can also be used as a departmental data mart within an enterprise, or as a companion test-and-development environment for a larger Teradata system. The appliance offers:
1. Integrated solution. Supports—in a single-cabinet, fixed-configuration, one-node system—all of the analytical and workload management features of the Teradata Database for applications requiring up to 6TB compressed user data capacity (based on a typical compression factor of 40%)
2. Availability. Provides RAID-1 disk storage and redundant hardware components, and supports optional uninterruptible power supplies

TERADATA EXTREME DATA APPLIANCE
Analysis of behavioral or movement data previously regarded as too large to be affordably used, e.g., e-commerce Web clicks and cell phone use, can be achieved with the Teradata Extreme Data Appliance. While high-volume, non-enterprise, possibly unstructured data is perceived as less valuable than traditional EDW data, its intense amounts of detailed information offer potentially enormous payoffs. As an adjunct data warehouse to the EDW, the system derives this value cost-effectively through:
1. Scalability. Combines the platform family technology with a density enhanced storage-array system of 50TB in compressed user data space for each node, enabling a total system data capacity of more than 50 petabytes
2. Analytic engine. Applies the full capabilities of the Teradata Database with four levels of workload management to provide the massive data set analysis solution for small numbers of users

Selecting a data warehouse architecture
By answering three questions, an organization can select a data warehouse architecture that fits its strategic and operational needs:
1. What is the business strategy? The organization must first consider how it will leverage a data warehouse to reach its corporate goals. The value of the data warehouse and the ROI dramatically increase as the company moves to the complete vision of a fully integrated and active enterprise environment. At the same time, specialized data warehouse applications used as independent data marts can be of an immediate, albeit lower, value. In the end, the value-driven ROI determines the level of investment in the data warehouse architecture.
2. What are the enterprise's plans? How the company plans to use its data warehouse as a tool to propel the organization toward success will determine the scope of scalability and functionality needed for the system. This also determines the importance of investment protection to the company. For instance, the need for high scalability would typically mean maximizing the value of the required sequential investments in the data warehouse.
3. What kinds of work will be done? The data warehouse must provide the right performance for an organization's data analysis types and workloads. As the mission-critical nature of the work and data increases, so does the need for high system availability.

Where the organization fits in its data warehouse evolution is based on various criteria, such as whether the system is used for strategic decision support or operational/active use. Other considerations are the concurrency and number of users, and the capability of workload management to meet service level requirements. Likewise, cross-functional subject areas queried and query complexity are factors in determining the best system fit.

What is learned through these questions will help a company determine the criteria for its basic data warehouse architecture needs and where it is positioned on the path to active data warehousing—the ultimate goal to enabling real-time decisions.

> Data warehouse appliances consist of processing, database, storage, inter-connection and software components that have been pre-packaged, pre-configured, and pre-optimized to accelerate core data warehouse functions such as multi-dimensional OLAP queries and bulk data loading.

> Teradata is widely acknowledged as the pioneering vendor in providing appliance-like data warehouse hardware/software solutions, having provided such solutions since the mid-1980s. Teradata continues to provide purpose-built, integrated solutions for data warehouses of all sizes.

> No data warehouse appliance pure play can offer a commercial platform as scalable—in database capacity, mixed-workload support, concurrent usage and fast data loading—or as flexible in deployment, optimization and administration features as data warehouse market leader Teradata, which has long had an appliance-like value proposition.

Source: "Appliance Power: Crunching Data Warehousing Workloads Faster And Cheaper Than Ever," April 2008 by James Kobielus, Senior Analyst, Forrester Research

Matchmaking
Based on the responses to the questions in the previous section, a company can establish criteria to identify the appropriate platform for its data warehouse. These basic criteria and their considerations are:

	Database functionality. Feature use increases as the organization moves from simple to active reporting.
	Data scalability. Data size expands as the warehouse becomes enterprise-wide.
	Query concurrency. User base and query volume rise as the scope grows to encompass the entire company.
	Cross-functional complexity. Data sources and applications increase as the environment moves to an active EDW.
	System availability. What is considered acceptable downtime in an entry-level data warehouse cannot be tolerated as the warehouse becomes operational.
	Workload management. For an active data warehouse, system resources must be managed for strategic and near real-time operational workloads.
	Investment protection. A small department system stays static while an EDW regularly grows; processor and storage capability are added but without costly system replacements.
	Active access. As the data warehouse becomes active, near-real time data loads and analysis are necessary.

The impact of each criterion and the extent to which each must be supported is determined by the enterprise scope, the scale of the data warehouse and its targeted operational/active level.

Figure 2, above, demonstrates how the impact of each criterion grows as the scope, scale and active use increase.

Family power
While the greatest value and competitive edge comes from achieving the Active Enterprise Intelligence vision, organizations are at different stages in their data warehousing experience. The Teradata platform family can accommodate this diversity with its purpose-built levels of business technology. T

Jim Dietz, platform marketing manager, has been with Teradata more than 14 years.