Saturday, 25 February 2012
Big Data Is Almost Here -- Using a Storage Hypervisor as an Agent of Change
Big Data – ultra large scale data storage and analysis – is the data center equivalent of that big rock, but it’s not arriving without warning and it doesn’t have to be an extinction-level event for IT professionals. To the contrary, it offers a unique opportunity to re-architect your storage management infrastructure in a way that makes it more adaptable in every respect and more easily aligned to business needs.
As profiled in the New York Times last Sunday, Big Data is transforming business, government, and education. One researcher reported that a study of 179 large companies showed that those adopting the data-driven decision-making that Big Data makes possible “achieved productivity gains that were 5 percent to 6 percent higher than other factors could explain.”
This shows that Big Data is more than just big. It’s restless, too, best used when hot. Let it cool off and you lose the situational awareness that can lead to big-time financial rewards. It’s not just a matter of storing a gazillion bytes – you can’t possibly store it all, so your retention policies have to change, and the need to widely share it as quickly as possible means your networking strategies have to change as well.
Fortunately, there’s a fundamental storage technology that can be a big help in adapting to Big Data: the storage hypervisor. Even better, the benefits of this software layer, which insulates you from all the hardware variables that Big Data can throw your way, kick in long before Big Data arrives. A storage hypervisor is an agent of change: you get the pay-off today and a future-proof storage infrastructure.
A storage hypervisor enables you to pool resources, automatically allocate space and direct traffic to the optimal tier, cache data near applications for higher performance, and manage it all centrally.
Resource pooling has the most immediate impact, because you can aggregate all of your storage capacity, without regard for brand, model, or interface, and easily reclaim unused space. Looking forward, this capability is key to integrating on-premise storage with cloud storage – a necessity to keep from getting squashed by Big Data.
The automation offered by a storage hypervisor gives you just-in-time storage allocation for highly efficient use of disk space, and the ability to dynamically direct workloads to the right storage resource (auto-tiering), based either on access frequency or business rules, so that the hottest data gets the most attention. With auto-tiering and bridging capabilities like Cloud gateways, the right storage resource includes not only disk devices, but solid state disks or flash memory devices for performance, or Cloud storage providers for virtually unlimited capacity. This makes it easy to balance data value and the need for speed against price/capacity constraints, something that Big Data is going to make ever more necessary.
A storage hypervisor can also cache data in main memory for rapid retrieval and fast updates. This turbocharges native disk array performance, especially if it’s combined with self-tuning algorithms. The result is that even off-premises storage can look local – again, a big “win” for Big Data.
Finally, with Big Data, your storage infrastructure is only going to get bigger, so centralized management of all your storage resources is a must-have. It gives you the equivalent of a universal remote for storage, no matter what it is or where it’s located, and is key to managing the mirroring and replication needed for high-availability, disaster-proof storage.
The fallout from Big Data is going to transform business computing at every level, so if you don’t want to end up a data dinosaur, now’s the time to transform your infrastructure with a storage hypervisor. A good place to start is Jon Toigo’s Storage Virtualization for Rock Stars series, starting with Hitting the Perfect Chord in Storage Efficiency, which will give you a good overview of how a storage hypervisor can help you increase engineering, operational, and financial efficiency.
[Photo Source: http://commons.wikimedia.org/wiki/File:Impact_event.jpg]