Making Data Warehouses Smarter
The ability to transform data into information and knowledge is at the center of business success. Decision-makers at every level of an organization require pertinent information to do their jobs well. An uncertain economic climate makes useful business intelligence crucial.
Effectively managing data and extracting value from it is a constant challenge for business and IT executives. Nowhere is the task more challenging than in today’s massive data warehouses. The relentless accumulation of real-time data, not to mention the size and complexity of these information repositories, can stretch organizations to their limit (or beyond it).
As a result, many companies are searching for a better way to leverage their data. Some are moving toward much higher performance analytic database systems and data warehouse appliances. These solutions typically consist of an integrated set of servers, storage, operating system and database software that’s specifically designed for complex analysis, data mining and reporting.
A key player in the analytic data warehouse space is San Diego and Cupertino, Calif.-based ParAccel. The company’s flagship ParAccel Analytic Database (PADB) software leverages the benefits of high performance and existing enterprise ecosystems to take analysis to record levels. Businesses looking to improve their analytic requirements have taken note.
Going with the Data Flow
The idea of storing data in a huge repository isn’t new, but it has evolved over the years. Along the way, storage requirements have swelled from gigabytes to terabytes and petabytes. Even with the development of complex and robust architectures that offer high scalability, the ability to execute tasks has begun to bog down.
In fact, Gartner noted earlier this year that its clients increasingly report data warehouse performance problems during queries. It estimated that nearly 70 percent of them are experiencing performance-constrained issues of various types.
“Many organizations are pushing performance to an extreme,” says Dipesh Patel, ParAccel’s senior manager of product marketing. “Completing a query or answering questions can take hours or days using traditional database processes.”
Data Warehouse Solutions
Enter high-performance analytic database solutions and data warehouse appliances, which can be modular, making them easy to install and scale as needed. In some cases, they can be configured as a cluster, with the database software preinstalled on the appropriate servers and storage devices.
Founded in 2005, ParAccel is a thriving participant in this market segment. The company’s high performance analytic database system, PADB, aims to take data analysis to a higher level and, in the process, reduce processing time and human involvement. “The goal is to bring greater performance to the data warehouse,” Patel says.
PADB spots patterns, relationships and correlations through complex and iterative analysis. It also offers operational agility, including an ability to explore any hypothesis regardless of the business questions.
In some cases, Patel says, PADB reduces the time it takes to complete a task by 95 percent or more. But the benefit extends beyond time and expense. “The ability to drill down to an unexpected answer that nobody has tapped into is extraordinarily valuable,” he says. “Organizations have an opportunity to gain a competitive advantage by leveraging data more efficiently.”
Businesses Take Notice
The analytic database warehouse is designed for a wide range of industries and situations. OfficeMax, for example, has turned to ParAccel to ratchet up its analytics and to drive more efficient buying decisions. Improved analysis, data mining and reporting allow the electronics and office products retailer to conduct more frequent market analysis to better understand customer behavior. The company can do this even on an intraday basis, if needed.
OfficeMax decision-makers come away with a deeper understanding of the product types in which customers are interested. The solution also gives them valuable insight into which products customers will add to their shopping carts.
“The answers they get allow OfficeMax executives to figure out how to position merchandise, bundle items and take advantage of product combinations,” Patel explains. “The strategy is designed to boost the number of products in shopping carts and, in the end, increase the retailer’s share-of-wallet.”
Another company embracing the analytic database concept is Provisio of Nashville, Tenn., which develops iTrials that automate and streamline the way drug trial candidates are identified, notified and enrolled.
The process includes leveraging an iTrials data warehouse — and its comprehensive longitudinal health histories of more than 70 million patients. Here patient-level data is combined to provide insight into patient populations and to identify factors that aid in protocol design, site selection and patient recruitment.
In the past, Provisio managed its database across several SQL Server clusters, Patel says. Pinpointing the relevant patient information was difficult and could take weeks to complete.
Today, Provisio uses PADB on a Linux platform to deliver patient information in columnar form. “The firm has consolidated 230,000 tables down to just 12,” Patel says. “Often, answers are available in seconds or minutes.”
The fact that PADB doesn’t require specific hardware is attractive to many companies. Notably, the software runs on an array of server and storage systems from such vendors as HP, NetApp and Oracle (Sun Microsystems).
“This makes it possible to install the software using existing server and storage standards in the data center,” says Michael Weir, senior director of marketing at ParAccel. “Companies are able to approach things in a less disruptive way and configure the environment to their specific price and performance requirements.”
Server Virtualization Connection
Deploying ParAccel’s PADB in a virtual environment could allow a business to achieve record levels of analytic performance for its private cloud. Such was the case when the firm established an industry benchmark for analytic processing performance this past April.
To learn about PADB’s record-setting performance, go to BizTechMagazine.com/
PADB410.
“VMware has revolutionized IT with virtualization solutions that enable efficient hardware utilization, reduced administrative costs and significant energy savings,” says Barry Zane, CTO and founder of ParAccel. With it, “we’ve shown the industry that the powerful compression and processing capabilities of PADB can be virtualized to deliver exceptional performance and savings. We have demonstrated new levels of analytic performance and IT efficiency in standard deployments.”
Make no mistake, the data warehousing landscape is changing. As a growing number of companies turn to analytics to slice and dice data in new ways, ParAccel will be looking for faster and more efficient ways to get the job done.
“Managing data more effectively and gaining insight into patterns, trends and behavior is critically important,” Patel says. “In today’s business environment, organizations that put data to maximum use gain a distinct competitive advantage.”
