The Challenge of File-based Data Storage

An Ask the Expert feature with Chris Santilli of COPAN Systems

Chris Santilli, CTO at COPAN Systems, explores the need for a radically different approach to solving the challenge of file-based storage growth.

“It’s this huge amount of persistent data that is most often stored unnecessarily on expensive fast spinning, hot disk arrays that is the heart of the problem.”
-Chris Santilli, CTO at COPAN Systems

The challenge of managing the cost and complexity of file-based data growth is one of the top five most difficult issues facing Global 5000 companies today. The growth and proliferation of file servers and Network Attached Storage (NAS) solutions have dramatically improved the performance of file system access but have created other problems around management and back up. Traditional NAS and server platforms are optimized and designed for transactional file serving performance and not for scalable long-term data storage.

So, what are we storing that’s causing us the biggest problems? With skyrocketing growth rates forecast, a CXO needs to understand what their data centers are being asked to store and why. There are two primary categories of data: dynamic and persistent. Dynamic data is fluid and changes regularly, today’s activity on your bank account would be a good example of dynamic data, whereas persistent data doesn’t change, it is static. A copy of your bank statement from June 2007 is a good example of such data – your bank will still need to keep a copy and have online access to it, however the data is persistent and the likelihood of needing to access it is very low. Most dynamic data turns persistent within 30 days. This is the heart of the problem – it’s this huge amount of persistent data that is most often stored unnecessarily on expensive fast spinning, hot disk arrays.

Archiving

File system archiving has been rapidly emerging as one of the primary functions that customers need to implement. According to SNIA, the average amount of persistent files on storage systems is approximately 80%. This means that 80% of files can be moved to a storage platform that is less expensive and easier to manage. The challenge has been identifying the data to move and having the right platform to ensure reliable long-term storage of digital file assets.

COPAN Systems have partnered with Quantum to enable the efficient long-term storage of file based persistent data. COPAN Systems’ purpose-built persistent data platform incorporates scale, density, protection and fast data recovery. Quantum StorNext data management software enables customers to store file data across multiple storage infrastructure types, yet file systems appear as normal, directly accessible by archive applications.

COPAN Systems’ Enterprise MAID (Massive Array of Idle Disks) technology is purpose-built to hold large stores of persistent data (up to 896TBs in a single frame with a footprint of less than 1m 2). It operates on the principle that, for long-term storage, A system’s disks don’t need to be kept constantly spinning. Instead, a maximum of 25% of the drives are powered at any one time, the rest are simply turned off.

And the results are simple, if the disks are not spinning, they are not using power or generating heat – this means they do not need to be cooled. If the disks are not spinning, then they are not vibrating which means they can be packed tightly together and offer considerable savings in footprint through this increased density. And because the disks spend less time ‘working’, their drive life increases by up to six times and they experience less downtime. This increased reliability, of course, means better data protection. There’s also the added benefit of having to ‘provision’ significantly less power and cooling infrastructure than the 100% spinning disk arrays require.

COPAN provides industry leading density, scalability and reliability for persistent data. Combining this with Quantum StorNext, you have the industry’s first complete file archive solution designed to handle your persistent file management, both today and in the future.