The changing face of the enterprise storage market [Q&A]
The growth of cloud usage, increasing volumes of information and a switch towards software defined systems based on commodity hardware have all had an impact on the storage market.
About a year ago we spoke to Mark Lewis, Chairman and CEO of storage specialist Formation Data Systems to get his views on the future of enterprise storage. This week we caught up with Mark again to find out how much the market has changed in just a short time.
BN: What have been the big changes in the storage industry over the past year?
ML: It's been an interesting year and in many ways it's been a validation of many of the things we talked about last time. The Dell EMC merger validates that we're moving away from the old system of doing things.
Storage and capacities are growing at some 40 percent a year, but revenues are in decline. Increasingly we're moving towards a world where software defined storage will replace traditional physical arrays. Just as VMware decoupled the computing environment from the hardware, software is separating the hardware from network and storage services. There's no doubt that this is how things are going, we're moving to mainstream market adoption, the only doubt surrounds timing. Most of the major cloud storage players like Google and Amazon are already software defined and increasing numbers of large and small companies are now looking seriously at the environment and beginning to test the waters.
BN: What's driving the change, is it bigger volumes of data or the need to make cost savings?
ML: At the core it's about being able to store more for less. Software defined storage offers a lot of good things, but it has to offer better economics than its predecessors. There’s a secondary set of benefits in terms of the ability to scale more dynamically, the agility and flexibility of the systems, not being locked into data silos, and simply being able to do things more quickly.
The software/hardware separation means enterprises don't feel tied so much to a single vendor. Customers can mix vendors giving them the ability to create a more competitive environment. They can choose to keep their hardware for longer or to switch platform vendors when it suits them.
BN: How much of this is coming from big data and IoT devices?
ML: It's clearly a big part of it. Companies don't want to raise their budgets but they're getting more IoT and analytics data, and they're dealing in richer media like video and in higher definition. When a business looks at its storage capacity and decides it needs an increase, it then looks at the price it's paying for its existing systems and decides it can't afford it. Economic necessity therefore drives it to look at alternatives, although other factors like simplicity come into it too.
BN: Is the storage market moving to smaller players or do things like the Dell EMC merger indicate a grouping together into bigger businesses?
ML: A recent Gartner report suggested that vendor size shouldn't be a factor in buying storage. There are so many technology startups that just because a company is big doesn’t mean it’s necessarily delivering good value or good investment protection.
Consolidation like the Dell EMC merger is actually because the market for legacy solutions is shrinking. Having said that it's a move that makes complete sense, because the market they're in will be about scale and providing the lowest cost for industry standard, commodity equipment. I don't think it means they will scale their software business though, in fact Dell is selling off all the software businesses it can.
We're going through a time when we need to scale differently. What I mean by that is that for a long time storage has been in vertical silos, so companies like Oracle would try to sell you a database plus the data management and storage array to go with it to lock you into a complete silo.
But the most efficient technology markets now have a horizontal model, using industry standard servers on which you overlay one or more hypervisor environments and operating systems, each of which is separable. So today if someone is trying to sell you an application or a database, you want that functionality to be portable so you can use it on your existing infrastructure. Like when you buy cloud storage from Amazon, you don't buy a database from them too, you just use their cloud storage servers and you want the flexibility to be able to move elsewhere.
BN: So we're seeing a shift in value from hardware away to software?
ML: If you can differentiate the hardware with useful features then you can still win by selling that. What you won't be able to do is sell someone the hardware because they've bought your software. If someone sells you software that will only run on proprietary hardware they don't have to make the hardware all that good because you'll have to buy it anyway. When software can run anywhere you don't have that problem. The same is happening with storage, the hardware vendors can compete on who can build the best enclosures, then we will compete in providing a virtualization layer to handle the data. But we don't build databases, we just make sure they will run well on our storage environment.
BN: What will be the major challenges in the storage market over the next year?
ML: It's going to be challenging for both customers and for vendors. Storage has gone longer than almost any other IT market since it's had a real disruption. It may have had the odd tremor but there's no disruption, for example EMC became market leader around 20 years ago and it's still the top brand.
Software defined storage means that customers will have to become comfortable with evaluating the benefits of different storage products or they're going to fall behind. Companies that don't start looking at this in the next year are going to be lagging behind, we're well past the early adoption stage. If you can't buy the extra storage needed to store your analytics data then your competition will move ahead.
There will be competitors to the major cloud storage vendors like Amazon, Google and Azure, no one can have that much market share and think that they can get away with no competition. Startups will eat into that market with niche offerings. But the real competition won't come from IBM or Dell or other big players, it will come from people we've never even heard of yet. We saw this with Amazon itself, when it first entered the cloud market everyone thought it was just a bookseller and it couldn't possibly compete. But it succeeded because it wasn’t burdened by existing ways of doing things, it did something truly better.