The Storage Array Killing Fields

“Best of breed” is the rallying cry blaring out from non-aligned third-party vendors as integrated IT stacks form from the vBlock trio, HP, Dell, HDS and IBM. The integrated stack people say open industry standards are lifeboats for users in if they want to abandon an integrated stack ship or make it work with existing networking or storage or compute gear.

So they are too but, in the long run, people don’t use lifeboats unless the ship sinks. There are more forms of lock-in than proprietary technology. Customers facing messages from the integrated stack people that say, yes, we support open standards, but if you use our stack components together you get more cost-savings and data centre efficiency savings than if you use third-party components for any part of the stack. Oh, and by the way you can order our stacks with one SKU and have one support contract cheaper than a part-component one with a separate SKU and support contract for the third-party component.

And, by the way again, our stack management software gives you a single pane of glass and you will need one more for each third-party component.

Why shouldn’t customers buy this story big time? Let’s use the car analogy. I buy a BMW and it has seats, probably made by a contractor who supplies BMW and Mercedes and Audi and whoever. I don’t know who they are and don’t care; I’m buying a BMW or a Mercedes.

There is one area where seat branding does matter; sports cars where I’d like Recaro seats, and some sports car manufacturers do use this best of breed Recaro seat branding. The thing is there are few seat manufacturers now that are independent in contract supply terms from the integrated vehicle stack manufacturers.

Who cares to bet that today’s list of non-IT system vendor, small and large storage array manufacturers: 3PAR; Atrato; Buffalo; Compellent; Data Direct; Dot Hill; Infortrend; Isilon; EMC; HDS; Huawei; LaCie; LSI; NEC; NetApp; Nexsan, Panasas; Pillar Data; Promise; Xiotech; Xyratex; etc., won’t be whittled down greatly as, or if, customers buy the integrated stack story?

In fact, if integrated IT stacks do take off, and my intuition says they are going to become pervasive in the next ten years, then best of breed could mean just that, the single best of breed with everyone else going to the wall or becoming a hidden contract component manufacturer for the integrated IT stack vendors.

That means guard your system vendor OEM and reseller deals guys. Do your best to turn your reseller deals into OEM deals, because if you aren’t the absolute and clear number one in your field and don’t have OEM deals with an integrated stack vendor, or become a fixed part of an IT stack vendor by merger or acquisition, you are going to go to the wall; become someone’s lunch; get crucified.

In ten years time which of the vendors in the list above will still exist outside of OEM deals?

If we say that just one will, the absolute best of breed then, I’m sorry, but based on storage mode coverage and revenues and channel strength and so forth it is NetApp.

The others will lose out in their ability to sell to general businesses and service providers. It won’t happen quickly, it will be a long and drawn out strangling of an industry but who will bet it won’t happen? Who buys special seats and special shock absorbers and special tyres for their car, van, truck or coach these days? Only the enthusiast, not the everyday joe; joe company, joe public sector body, joe taxi company, joe limo company, or joe trucking company.

In what respect does a storage array differ from a car seat or a shock absorber when the thing it is going to be fitted to is an integrated manufactured end-user product, be it IT stack or a vehicle?

I mean, in ten or fifteen years time, who cares, who really cares if the majority of the storage vendors above exist? Is there a remembrance society for lost seat manufacturers, vanished shock absorber suppliers?

Who mourns Auspex, OnStor and other vanished array suppliers now?

The storage array vendors, whether they like it or not, are going to be engaged in a multi-year, titanic struggle for survival. “Are going to be” … they already are. Oracle has Sun, with Pillar Data in the wings. HP has StorageWorks. Dell has EqualLogic. IBM has its DS whatevers and XIV and SONAS. Fujitu has its Eternus stuff, including the old FSC storage. HDS is part of Hitachi which has servers. NEC has HYDRAstore and other storage stuff in Japan. Huawei has is own storage. Acer needs some storage supplier tied to its servers if it is going to become an integrated IT stack vendor.

Everyone else is out in varying degrees of cold. Dot Hill, LSI and Xyratex are the three lottery winners. They have OEM supply contracts as their core business model. Guard those contracts boys; they are gold nuggets in a world that’s going to be full of gold rush miner vendors looking for their claim.

And the best way they could get it, the best way they could get on the OEM storage array supply and survival gravy train? Become a claim-jumper and buy one of the now, oh so very valuable threesome: Dot Hill; LSI; or Xyratex. These three have the keys to the OEM supply door. It’s easier to buy a key holder than design, build and establish your own OEM supply door.

The storage array vendor killing fields; that’s what I’m blogging about in this apocalyptic blog. Am I going over the top? Let’s talk in ten years time and, meanwhile, when you slide into your Chevy Suburban, your Ford Crown Victoria, your Merc E Class, your VW Passat, and so forth, ask yourself this: “Who made the seats?” And: “Do I care? I mean, really?”


About General_techy

I'm a storage industry journalist and this blog is my place to put articles not published by the mainstream storage media.
This entry was posted in Uncategorized. Bookmark the permalink.

One Response to The Storage Array Killing Fields

  1. Portencia Monacle says:

    A terrific analogy. Over the years the special, simplifying power of the car analogy has made countless points seem true. But on this occasion, one wonders about its applicability. Consider the core motivation behind so many of the recent mash-ups such as HP & EDS, Dell & Perot, Oracle & Sun, VCE, etc… These were necessitated by a primary strategic objective: to command a greater share the customer’s budget and share less of it with others. In other words, a land grab. Of course, that crude justification doesn’t sound like value-add in front of customers, the press, or analysts. So what do you do… you tie your move into the latest buzz word: “cloud computing.” You assert that cloud computing — which is so different, or so complex, or so _____ (fill in the blank) — requires that it be delivered in a pre-integrated stack. And to the extent you have weak links in your stack, you assert that the value of pre-integration and one-vendor support is so great — somehow magically, because we are talking about cloud computing now — that differences in the components of the stack become trivial. And with a little help from your friends — using a suddenly relevant car analogy — you can make this “new and preferred” way of IT buying almost seem inevitable.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s