Like all previous generations of information technology, the latest breakthrough has been undermined by the very people who are supposed to be selling it. Storage virtualisation has a simple aim and one that most IT buyers would enthusiastically endorse. It aims to reduce complexity, represented to end-users by separating the physical media on which data is stored from the server operating system. In other words, users don't need to know which device their information is stored on - the operating system worries about that - so they can spend less time getting their information and more time doing their jobs.
This is a simple objective, and one that most people would understand and applaud, although it doesn't help when the concept is given a silly name like 'storage virtualisation'. But despite the confusion that its name will cause, storage virtualisation will be the next boom area of IT, and IDC predict that the market will be worth $53 billion by 2004.
An evolutionary process
There are already a number of key players in this market. First there's the traditional storage giants like Compaq, Dell, EMC, Hitachi, Network Appliance, Sun and IBM. Then there's the newer, more specialised players, such as Veritas, Auspex, Brocade, CacheFlow and Storage Networks.
Since there are several stages of evolution to be addressed in this market, there are any number of aggressive newcomers that might be worth considering in the next few years.
Companies will target areas such as multi-protocol switches, SAN software, iSCSI products, Internet Acceleration and caching and massively scalable storage systems with their own innovations.
So in the coming years, the storage virtualisation market should offer plenty of opportunities for picking new vendors, identifying new markets, creating new services and, most importantly, making lots of money.
All the major players and the new upstarts, then, are anxiously trying to devise strategies for this important new market. It will be imperative that they send out clear, unambiguous messages about their plans for helping users to improve their IT storage management systems by many degrees. Let's hope they do this more effectively than whoever it was that came up with the name for this new concept.
Some degree of respectability will be established when the first referenced customers of storage virtualisation begin to emerge at the end of 2002. Even then, there will continue to be doubts over the future direction of this technology.
As ever, there seems to be two schools of thought over the direction of storage virtualisation. Some insist that software companies will be well placed if they can promote the concept of interoperability with, and independent from, third party hardware. They will be regarded as translators, providing the openness and flexibility that affords users choice over their hardware suppliers.
On the other hand, there are those who insist that, being purely practical, storage virtualisation will only ever be accomplished by using a single vendor, and that ultimately the company that will win in this market will be the one that offers the complete range of hardware and software. The resellers that will profit the most will be those that back the vendor most likely to triumph.
In other words, openness is something all vendors will pay lip service to (just as the networking vendors did) but the vendors that triumph (as Cisco did) will be those that can assimilate the most complete spectrum of products, so buyers only have to go to one source.
Before then, there are the usual standards hurdles to be cleared. So obviously there are a lot of challenges that have to be met before resellers can start issuing their huge corporate clients with invoices for services rendered around storage virtualisation - it could be years before the market takes off.
Help for resellers
It's too early to say how the storage evolution market will evolve, but it's illuminating to examine which types of resellers the major players are looking to attract.
Bob Iacono, marketing vice president at Auspex Systems, admits there's not too much evidence of a market at the moment. "All the evidence is anecdotal to date," he says. "Lots of hype, lots of different definitions. Most users don't understand what it will do for them. The pioneers in the space, FalconStor and DataCore, don't have much revenue as of yet, although they are attracting early partnerships."
In the meantime, Auspex is planning to expand its market by recruiting resellers. Iacono anticipates that the large Unix/SAN resellers will be the ones that are most immediately appealing. This is because they have the ability to integrate with existing solutions and to install to a customer's benefit, he argues.
But why should resellers commit to selling a technology that's clearly not mature? There are too many conflicting ideas about standards, and many of the different vendors have different definitions of what storage virtualisation is. So what is Auspex doing to ease them into this market?
"The resellers want some market pull and the ability to make a profit," says Iacono. "At the moment neither are there. From there they will require system administration tools for ease of configuration and management. They will look for standards to install across a variety of platforms. We do offer co-op moneys. We're planning on doing a couple of major storage focused shows in Q1 and Q2, with the major theme being SAN/NAS cohabitation and interoperability."
Adapting to virtualisation
While Auspex might represent the views of the recent entrants to the storage business, it is worth considering the traditionalists. Tandberg Data, for example, has a legacy of helping businesses manage their information. Sales manager Steve Smith says that as tape drive applications become more complex and automated, resellers in this space will be able to gear up to offer storage virtualisation.
"We're addressing the market by looking for resellers who understand the tape storage market and the various new applications for on-line, near-line and off-line tape storage," says Smith. "The tape storage market is evolving into a new era. It's no longer about backup."
He adds that the most important consideration in this market at the moment is not so much the vendor's background (it's too early to judge which standards will come out on top) but the level of assistance that the vendor offers. "The highest priority at the moment is sales and technical training, closely followed by sales leads and co-op marketing," he explains. "We can provide all of those and can demonstrate the potential to add incremental business and a basic knowledge of the tape market."
Explaining the concept
Glenn Hintze, storage business development manager at IBM, also argues that the companies that have been around in the storage market for longest are best adapted to meet the needs of virtualisation. "The idea of virtualising resources in computer system design is an old one; we have had virtual memory, virtual systems and virtual devices of all kinds for years," says Hintze.
"In fact, we have had several types of virtual storage for some time. To understand what is new in this area, we need to explain the basic concept of virtual storage, how storage networking systems use storage and the opportunities to improve these systems with virtual storage."
The basic concept of virtualisation is a simple one, he says. Hardware and software interfaces are defined to allow various components to be assembled into a working system. Once interfaces to a component have become set, designers are strongly motivated to preserve them as they improve the component.
Often designers will produce a new version of the component that preserves the external interfaces but is totally different in implementation. This new version, that closely mimics the interface behaviour, presents a view of an idealised component, a virtual component, that isn't really there.
"Sometimes the advantage of the virtual device is so compelling that the virtual device becomes the standard for system design," says Hintze. "Virtual storage is collections of disk images or file systems that are presented to applications or middleware, but are re-mapped in some significant way before data reaches physical storage."
In the meantime, there's plenty of scope for resellers to make money - as long as they don't make the mistake of giving their consultancy away free in lieu of making margins on sales.
Mel Taylor, marketing director at Fujitsu Siemens Computers, says there's plenty of areas to make money. "The revenue at the moment comes from technical integration," he says.
"Get the virtual storage system into your own portfolio and business proposition. Then you should offer professional services, like design and implementation." The most important consideration is that storage virtualisation is mainly based on SANs, so these skills should be acquired as a foundation, concludes Taylor.
Not so simple
The first virtualisation products are gradually coming on the market now. Despite all the aggressive newcomers in this market, a lot of analysts seem to think the established players have the easier task in capturing this market. After all, it's a lot easier to adapt an established product portfolio for existing customers than to go out and establish a new customer base.
Compaq is aggressively pursuing a SAN Systems Integrator programme to attract more resellers into its fold, and feedback from resellers is generally favourable.
"It's an uncertain market, so we'd much rather go with an established player than some new unknown prospect," says one reseller. "In the current economic climate, it's a good bet they'll go under or get bought."
Not an argument that Geoff Barrall, founder and CTO of BlueArc, one of the new storage specialists, would agree with. "Virtualisation means something different to everyone," he says. "Broadly, it can be defined as any technology that makes something look simpler than it actually is."
If only there was a way of simplifying the challenge for some of the start-ups. The stance on the storage virtualisation market is that established players will see off the challenge from newcomers. And these are probably the vendors resellers should be getting into bed with.
This was first published in January 2002