I've been watching and waiting for years for a vendor to proclaim the demise of disk storage, and this week it finally happened. The vendor who made the bold statement is Sun, specifically its CEO and President Jonathan Schwartz.
I admit that Schwartz didn't use those exact words when making the announcement in his blog. In fact, he opens his post by saying, "I'm radically increasing Sun's focus on storage today", which you might take to mean exactly the opposite of what I'm saying, until you get to the next paragraph:
"I'm going to be combining our storage and server product teams to create a new converged group at Sun known simply as our 'systems' team. The systems team will focus on the evolution and convergence of computing, storage and networking systems."
That's it. Over. Finito. Sun won't have a storage engineering team separate from server engineering anymore. Does this mean Sun won't continue to sell storage arrays? Of course not; arrays will certainly stick around for a while, thanks to the need to absorb ever higher tides of digital data. But the days of developing storage solutions independently from (or, I'm tempted to say, in opposition to) application servers are gone — at least at Sun, but other vendors take note.
The convergence of systems and storage is going to change the datacentre landscape dramatically, so you might want to take a good look now, and maybe take some pictures. In time, a computer room filled with storage arrays will be as anachronistic as the "disk farm" of time-sharing days.
Schwartz has a point when he says that, looking at a blade system, it's hard to tell where the server part ends and where storage begins. In a future where blade systems are ubiquitous, the distinction between server and storage won't be easy to make.
If blade systems are indeed the future, where does this leave the so-called pure storage vendors such as EMC and Network Appliance, not to mention the hundred or so minor players? Frankly, I don't have a clue.
Depending on how aggressive blade systems vendors become, self-standing storage solutions could hang around for quite some time. Perhaps they'll never quite disappear, but shrink into a market niche of limited relevance.
And if you're thinking that blade systems are for Fortune 500 customers only, think again. Both HP and Sun are reducing the tonnage in entry-level models to make them more appealing to a broader market.
Other vendors will follow, if they haven't already.
Interestingly, even if consolidating servers, switches, host adapters and disk drives into all-in-one blade systems could be the strategy that wins the datacentre, one piece of the storage puzzle will always remain independent. Tape libraries, because of their very nature, will remain outside, as detached units. Wouldn't it be ironic if the tape drive, a device that has been declared dead so many times by so many experts, were to be the only survivor of the storage years?