As the first full week of 2011 draws to a close, a common theme is already starting to emerge from the vendors and enterprise users with which I am speaking. In their minds, the era of buying storage arrays in the form of point solutions to solve individual problems is coming to an end. But it is not because large enterprises have not gotten so much smarter. It is just that storage vendors are waking up to how their customers actually think and behave and are finally delivering storage solutions customized for them.
I don’t think it is any secret that for years analyst firms have been telling users that the majority of the cost associated with any storage array purchase is the management costs. Of course, did anyone actually listen to that advice?
Not that I can tell. When I look at the evidence in any large company for which I ever worked it always seemed to me that those individuals who rushed out to buy more storage arrays were the same ones who inevitably got promoted to higher positions within the company because it was perceived they could “get something done.”
Exactly what it was they were accomplishing was never entirely clear to me. Yes, they purchased new storage that provided more capacity, more performance or some combination thereof but when queried as to why existing storage array could not be better optimized or configured to handle the new workloads, their response was always nebulous.
In other words, they didn’t really have a clue why they were buying what they buying but it appeared to be the fastest, easiest way to solve the problem in front of them and they convinced their management team of the same thing so the purchase got approved and they got recognized for “getting something done.”
The only problem with this approach is that after 2, 3 or more years of doing this, having a collection of storage arrays sitting all over the data center floor dedicated to specific tasks with tons of excess, unused capacity was causing more problems (and costing more money) than it was solving. Further you had no way to manage all of these storage arrays or even connect one application server to another storage array to access its excess capacity without jumping through a bunch of hoops. So a lot of these individuals who made those knee jerk storage buying decisions ended up either finding employment elsewhere or just ended up unemployed.
Trying to go back and untangle their mess was really not an option either. The time and effort required to do so became so complex and time consuming that the best solution was to buy a really big storage array on which to consolidate all of the data residing on all of these smaller arrays.
That approach only sounded great until one got the quote for “the really big storage array.” Then suddenly having to manage all of those storage arrays sitting all of over the data center floor did not sound like such a bad option. Granted, you were still in the same pickle as before but at least your company was not broke and you still had your job, miserable though it may have been.
This is why I see large enterprises looking beyond traditional storage array solutions in 2011 and aggressively seeking out and deploying scale-out SAN and NAS storage architectures.
The reasons for this are simple. Organizations are not going to change. They are still in crisis mode. They are too fluid. They have less time than ever to focus on storage. They still don’t really have a long-term strategic storage initiative. And the people who “get things done” are still the most apt to get promoted even if their thinking on storage tends to be a bit short-sighted.
The good news is that by buying storage systems that fit into these emerging “scale-out SAN and NAS architectures” makes it irrelevant if your company operates in crisis mode, has no strategic plan, and is run by individuals who are convinced that storage decisions can be made in 15 minutes or less.
Using these emerging scale-out storage architectures, they can buy and/or upgrade their existing storage architectures or even virtualize existing storage systems with them and still keep their storage environment quasi-manageable. In other words, storage providers are finally waking up to the fact that they can’t change their customers; all they can do is adapt and re-architect their products to meet the way their customers actually act and behave. These new scale-out storage products are the end result.
Now do I believe these new scale-out SAN and NAS architectures are going to perfectly optimize data placement and manage data in the best manner possible? Probably not. But are these architectures an infinitely better match for how organizations truly act and behave and lead to organizations being more satisfied with their storage environments? Absolutely.