Reassessing Your Data Center Consolidation Project

Untitled Document

It seems it's always monsoon season in today's data centers. Regardless of the economic climate or any other factors, CIOs and data center managers can count on constant and enormous volumes of data of various types gushing into their servers from virtually all points across their enterprises.

At the same time, as fast as the data pours in, executives, financial analysts, marketers, sales, field service, customer support and other employees and business partners want direct, unmediated access to it. They want hands-on access to up-to-the-second data that they can explore on their own.

While there's a lot of goodness to be found in this environment of exponential data growth and the voracious desire to squeeze every ounce of competitive advantage out of that data, it does present a few pesky challenges.

When budgets were a bit more generous, IT simply added more capacity -- more servers, more data marts and data warehouses, even more data centers.

Those days, though, are gone. Today, IT must find ways to maximize capacity and performance using existing assets. In fact, many departments have been directed to find ways to reduce their data center footprints and operational costs.

Data center consolidation: high on CIO's to do lists

A number of surveys have found that at or near the top of almost every CIO's to do list is some sort of data center consolidation initiative. These initiatives are typically driven by the need to reduce server, storage and application sprawl and their accompanying costs. These projects are also intended to improve business processes, increase workforce effectiveness, and empower business users with the comprehensive information they need to target new customers, grow existing customer relationships, create new products or services, and expand into new markets and geographies.

There's also the matter of being able to comply more quickly and accurately with regulatory and financial reporting requirements.

And not to be forgotten, there's the growing trend among organizations to take steps to be "green" by reducing power consumption (and the attendant costs) and doing their part to address the climate change crisis.

Information explorers: organizations' new secret weapons

Business users are no longer passive data consumers, report readers or simple question askers. Armed with a multitude of sophisticated, user-friendly desktop tools and with increasing authority to reach directly into enterprise data stores, business users have become very active and intensive information explorers.

While it's still absolutely mission-critical for operational systems to run reliably while handling millions of transactions per second, organizations' real competitive advantage now comes from their ability to explore and analyze enterprise data. These information explorers are gaining the insights and finding the seeds of innovation that are critical to achieving and maintaining leadership positions in their respective markets.

It's for this reason that you should view your data center consolidation project as a golden opportunity to maximize IT's contribution to the corporate bottom line, rather than as some form of drudgery or punishment. View your project as a perfect vehicle for innovation that will allow you to increase the efficiency and cost-effectiveness of both your operational and analytical systems.

Match the platform to the task at hand

In general, it's a good thing to have all of your ducks in a row. In the case of your data, however, that old adage is faulty. Instead, develop your project plan with the knowledge in mind that traditional, row-based operational systems, no matter how highly tuned and optimized, are designed to process transactions -- rapidly, reliably, accurately and securely. They are not designed to support data, user or query-intensive reporting and analytics workloads in an economic way.

Each of these business-critical functions -- operations and analysis -- should live on a dedicated platform that's been optimized for that specific business activity. Organizations that rely solely on traditional, row-based RDBMSs to support these very different functions will find themselves maddeningly frustrated by the cost and complexity of providing high performance for transactional and analytical workloads. Unable to meet their organizations' ever-increasing expectations for either of the workloads efficiently, they ultimately may find themselves unable to compete, let alone dominate.

Instead, determine which of your business requirements are operational and which are analytical. Then begin planning to employ appropriately optimized database platforms for each. The chart below provides a starting point to help you determine what data are best handled by what types of DBMSs.

Count on columns for analytics

In today's business environment, operational excellence is not enough to propel your business forward. Instead, your ability to execute your strategies and achieve your objectives is determined by the quality of the decisions and predictions you are able to make based on deep analysis of your current and historical enterprise data.

When the task at hand is analytics, a column-based analytics server provides an efficient, cost-effective and scalable solution. Column-based analytics servers provide better performance at lower cost and with reduced hardware requirements than traditional RDBMSs. They are designed to support a large number of users concurrently running ad-hoc queries, giving the highest priority to query performance, the next highest priority to the speed at which bulk data update can be accomplished and a much lower priority to the performance of small data update.

Information explorers need analytics servers designed from the start to optimize advanced analytic activities including ad hoc queries, extensive data mining, forecasting, predictive modeling and optimization techniques intended to drive strategic and operational decisions and actions. Information explorers' goals typically focus on gaining better understandings of customers and competitors, identifying fraud, identifying and reducing areas of risk, optimizing supply chain performance, maximizing marketing program results, locating revenue leaks and more.

Ask yourself a few questions

It's unlikely that there will ever be a data drought. To the contrary, we are in the midst of an unprecedented and accelerating data explosion.

If you're still not sure that your organization would benefit from column-based analytics, ask yourself a few questions:

  • Do you need to run complex queries on huge volumes of detailed, real-time and historical data to make reliable predictions?
  • Is your current computing environment providing you with the answers and insights you require as rapidly as you need to outperform your competitors?
  • Is compliance with regulatory mandates becoming an increasing burden?
  • Are you experiencing a decline in your operational RDBMS performance?
  • Do you need to realize a better ROI on your technology investments?

If you answer yes to any of these questions, you owe it to yourself and your organization to explore the potential benefits of implementing a column-based analytics server as part of your data center consolidation project.

About the Author

Sam Friedlander is Director of Global Business Intelligence for Sybase, Inc. He attended the Wharton School of Business and has been a practitioner of Data Warehousing and Business Intelligence for 13 years.

More by Sam Friedlander