Financial Services Firm Shrinks Secondary Storage

For many years, secondary storage has continued to take second place to primary storage tiers in enterprise storage planning. But the growing popularity of techniques like data de-duplication and the emergence of "tiered value storage" are causing enterprises to take a look at the volumes of data that are not critical to their daily processing, but that should nonetheless be evaluated for archiving and then stored.

Wisely managing exponential data growth is one impetus for the growing enterprise focus on secondary storage, and on strategies that tier data into primary or secondary storage based on the data's perceived value (e.g., is the data regularly accessed or not). Other economic pressures on enterprises include shrinking IT budgets, and directives from executive management to do more with less. In the face of these pressures and also mounting workloads, corporate IT is looking to do whatever it can with existing resources -- in storage and elsewhere -- to protect staff headcounts. This means evaluating data that routinely used to just be backed up without question -- and getting rid of data that is duplicated or antiquated in the hope that new storage media purchases can be avoided.

"There is no doubt that companies like us are starting to see momentum in sales growth," says Tom Cook, president and CEO of Permabit, a supplier of secondary storage solutions and technologies that include data de-duplication, snapshots, data protection and archiving. "In a tough economy, the moves that enterprises are starting to make are all about money -- and what they can save. Companies like ours have redefined their cost models, driving down the cost to implement. We are also able to deliver efficient management for secondary data at attractive price points."

One case in point is Penson Financial Services, Inc. (PFSI), a financial services firm that provides processing services for the execution, clearing, custody and settlement of financial transactions. Financial regulatory and litigation support requirements mandate that data be stored for long periods of time, so Penson found itself adding disk drives and archiving data to DVD and optical WORM media. The process became problematic when the number of DVDs and optical platters exceeded the capacity of the jukeboxes. This storage environment was neither scalable, cost-effective nor labor efficient -- and locating data was time consuming.

"When we needed to comply with an order to produce historic documents and email, senior IT staff would disappear for days or even weeks to complete the request," says Matt Pittman, vice president of MIS Infrastructure for Penson. The company moved to a tiered value storage system provided by Permabit that Pittman said reduced document request work from weeks to minutes. Plus, "we achieved 100 percent [return on investment] in months." he says.

"This is not an isolated occurrence," says Permabit's Cook. "If you had asked me about tiered value storage eighteen months ago, I would have told you we were doing 'missionary work' out in the field, and that we were finding that customers were doing limited archiving, and that whatever they were doing was centered around regularity requirements and email, and that otherwise they were just doing standard backups of all of their data. But with the cost barriers to entry dropping, enterprise interest has really taken off."

Just what can enterprises save by moving to a tiered value storage concept?

First, secondary data is efficiently sorted out from primary storage data that you are actively processing. This allows IT to focus its most intense management efforts on a much-reduced subset of mission-critical data. On the archived storage side of the equation, Cook says that the potential data storage savings can vary depending upon enterprise and application. "Using a technology like data de-duplication with archiving, we conservatively tell sites that they can save 2x-20x on the amount of storage required for their data," Cook says. The savings is found from the amount of data that gets stripped out in de-duplication, so that only one instance of data is stored.

"However, if you have a site with a high degree of virtualization and instances of data and systems, you could see an operating gain of 60 percent to 85 percent, because there is significant data replication," Cook says. This ultimately translates into a reduced need for storage media and servers, which in turn translates into energy, equipment footprint and floor space reductions and savings in the data center. These are the quantifiable costs that give IT hope that it can preserve headcount while better managing storage and other computing sources.

InformationWeek has published an in-depth report on cloud storage. Download the report here (registration required).