This article can also be found in the Premium Editorial Download "Storage magazine: Disaster recovery planning: 10 tips to help you avoid costly missteps."
Download it now to read this article plus other related content.
Storage professionals building out their infrastructures have more new storage technologies to choose from than ever before, but the choices are becoming more complex.
Back in early 2012, I predicted 2013 would be the year of storage performance innovation. Over the past year, a number of new storage technologies have debuted, often stealing the storage performance spotlight from traditional big storage array vendors. The technologies have run the gamut from server-based to network-based storage acceleration, as well as entirely new storage array approaches that tier and cache much more effectively -- often even outperforming the incumbent enterprise arrays.
For 2013, storage performance will stay center stage, but the competition for your data storage dollar will ramp up like we haven't seen in a long time. It's been a while since the big array vendors have had to compete head-to-head on performance, and with dozens of startups offering a challenge to array performance, the competition is bound to get serious.
I'm predicting we'll see an entirely new level of investment and innovation from incumbent vendors in these areas:
- Architectural improvements designed to enhance scalability and controller efficiency
- More elegant integrations of solid-state technologies
- Innovative integration with server-side caching technologies
When you look at the list your first thought might be, "Won't the big vendors just acquire the smaller, technically advanced companies and evolve their product offerings?" You may even think you can just stand on the sidelines while the vendors duke it out and see which technology is left standing or ends up owned by one of the big vendors.
But I don't think it will be quite that easy this time around. If everyone takes a wait-and-see attitude, we may end up with a stagnant data storage market and lots of business initiatives that are worse off for it. I think there will be fewer acquisitions of upstarts by mature vendors, and in turn there will be a lot of new vendors in the market to compete for your storage dollar. Acquisitions aren't always very easy; there are two particular challenges for big incumbent vendors looking to acquire technology. First, it's not easy to integrate entirely different architectures with existing technologies, and offering dual (or dueling) technologies can be confusing for users. Second, the major vendors likely to acquire upstart technologies have all made recent acquisitions and their dance cards are already pretty full.
Meanwhile, the storage performance pioneers are beginning to create an enormous capability gap versus traditional storage architectures and, consequently, they're starting to get the attention of many users in what's now a very big marketplace. This will, in turn, drive the growth and lasting presence of many more storage vendors in the market than we've ever seen before. The list of new challengers is incredibly long, represents many different offerings and includes relative newcomers like Fusion-io, LSI, Marvell, NexGen, Nimble, Nimbus, QLogic, Tegile, SanDisk, Violin and Whiptail, and even some vendors, like Nutanix and Simplivity, offering hyper-converged products. These innovators aren't all small startups, but it's a distinctly broader set of vendors than the big four or five storage vendors we're accustomed to, and any one of them may come to be a pivotal provider of storage services for your data center.
Leading or lagging, putting it together gets tougher
This time around, and especially in 2013, whether you're a leader or a laggard with storage technologies, buying decisions will get tougher. First, figuring out how to place a value on these different products can be a challenge. We're used to thinking in terms of $/GB, with rough expectations of performance for different types of storage, such as enterprise class versus midrange. If we shift to a performance focus, $/IOPS makes a lot more sense, but then we need to be able to characterize how much performance we need, and entirely new ways of talking about performance will be name of the game in 2013.
Architecting an ideal storage infrastructure and maintaining optimal performance through changes and additions looks like it may be harder than ever before. The storage infrastructure is moving to a multi-tiered model, which makes weaving together components such as devices that can cache, optimize and store data across arrays, appliances, networks and servers a tough task indeed.
But putting together a solid solution is likely to be more important to your business than ever before. Business agility often revolves around high-performance data. And when that kind of performance is needed, these technologies can transform the cost model of your storage infrastructure for the better. Whether you prefer the cutting edge or only the tried and true, 2013 is shaping up as a year in which you'll have more choices than ever.
About the author:
Jeff Boles is a senior analyst at Taneja Group.
This was first published in February 2013