Late last year I pondered how storage might look in five years’ time. I looked at the effects of server and desktop virtualisation on storage, noting that the needs of virtualisation had given shared storage a leg-up to near ubiquity among organisations but also that its shortcomings threatened to see it eclipsed by forms of storage that brought server and storage (with tiers, including flash) much closer together.
While that was an attempt at assessing the trends -- existing and nascent -- in storage technology, it’s arguable that it is an incomplete one. It was an “internal” view, if you like, and one that failed to take account of the wider forces, mainly economic, that will shape our industry. So, here I’d like to look at storage in a much wider landscape.
The big theme I want to examine here, the one I see as potentially shaping the future of storage, is concentration of ownership, of vendor consolidation. It’s a fact of economic life. As capitalism in general and as industries within it have arisen and matured, the ownership of enterprises has followed a familiar pattern that sees a proliferation of businesses that then tends towards an ever-smaller circle of ownership until a few big players are left.
We’ve seen it in all the major industries. Whereas in the 19th century every busy town with thriving industries had numerous local banks, now each country has a few that dominate, and many are global in reach. Whereas initially individual men with a well and some draught animals drew oil from the ground and moved it to town, now the international oil industry is dominated by fewer than 10 companies. Whereas at the start of the automotive industry every major town had a blacksmith putting motor vehicles together for local customers, we now have a small vendor consolidation of multinational car makers.
In storage the ecosystem of established giant vendors and innovative startups seems to exist in a state of dynamic stability. The big players scrap over percentage points of market share, while the startups innovate and ultimately gain the attention of the big boys.
We’ve seen it in storage too. Go and look at the Wikipedia page on “defunct hard disk manufacturers” and you’ll see a list of almost 100 companies that have tried their hand and failed and either gone bust or been absorbed by the five that now rule the market.
This consolidation of vendors is a fact of economic life because the bigger a firm is, the more it can make economies of scale and push down supply costs and wield power generally in the marketplace. It’s also true that bigger businesses can generally weather bad times better than small ones, and for this reason every economic recession since capitalism began has seen accelerated waves of industrial consolidation.
When it comes to the storage systems market, we also have concentration, although the character of the industry is distinct from others. We have some very large storage vendors that take the lion’s share of the market. The “big five” includes HP, EMC, NetApp, IBM and Dell. But, that said, there are a myriad of smaller players, ranging from those just outside the big five, such as HDS, Oracle and Fujitsu, through to the micro-vendors and startups.
In storage the ecosystem of established giant vendors and (usually) innovative startups seems to exist in a state of dynamic stability. The big players scrap over percentage points of market share, while the startups innovate and ultimately gain the attention of the big boys, at which point they are often bought by them and folded into the larger companies’ efforts to win market share.
Corporate IT systems are complex and extremely technical. They are vast and made of many parts from many vendors, as storage systems within them often are too. This makes storage unlike, for example, the oil industry. In the latter, there’s no scope for someone coming up with a new type of oil that some people might decide will help their cars run better. The big oil players dominate from exploration through transport and refining to sales, so there’s just no way in for an oil company startup. The customer simply gets the type of oil product they need. Despite buying it from one oil company, they probably don’t know who pumped and refined it, and they don’t need to.
In short, oil is a commodity, but IT and storage are not. But will that always be the case? Not necessarily, and cloud technology could be the agent that forces this change. Right now we’re used to an environment in which the IT vendor delivers the end product, ie, the storage or the IT provision. If the cloud takes off, processing and storage could become a service, delivered to the customer from giant remote processing and disk farms. It is surely only a matter of time before bandwidth and security are up to the task.
When that day arrives immense shock waves will go through the vendor community, as suggested by Chris Mellor in this recent column for SearchStorage.co.UK. Effectively, what would happen -- framing this in terms of consolidation and commoditisation -- is that much of the IT vendor community would no longer sell directly to the customer organisation. Instead, the cloud provider sector would become the predominant buyer of IT and storage products. Such a sector, arising anew, would become highly standardised, buying huge amounts of equipment that would allow it to deliver processing and storage to end users just like pumping gas.
In such a world, consolidation among big IT and storage vendors is only likely to increase. No longer would there be a populous user customer community and appetite for point products among customer organisations. The provision of IT will have moved upstream. It’s true that the oil industry has its share of small, often innovative consultancy, outsourcing and specialised technical businesses, but the end user never sees them. This would likely be the case too in a cloud-based IT world. As the big vendors dominated sales into huge cloud providers, the ecosystem of startups and specialists would thin out to those able to coexist in the new environment.
All of which throws up the question: What kind of storage market would such a scenario demand? It’s all very well to extrapolate on the likely direction of storage technology from today’s economic realities. But those are not tomorrow’s. We may look back on these last couple of decades in IT like we look back at the age when blacksmiths built the first cars or mule trains carried barrels of raw crude from the Pennsylvania hills.
Antony Adshead is bureau chief for SearchStorage.co.UK.
This was first published in March 2012