Back from Minneapolis

Someone grabbed me from sunny Florida, threw a heavy overcoat, scarf and gloves my way and said, “Go to Minneapolis to see the future of storage.”  The future of storage.  Hmm.

I got off the plane and was bustled into a car driven by Bruce Caswell, Director of Marketing Communications for Xiotech.  Xiotech?  That was the company that some Wall Street guy had once told me had to spend $100 milllion to make $50 million.  (Not true, by the way.  At least not for the last four years.)

In any case, I was anxious to meet up with Steve Sicola and maybe Ellen and/or Ritchie Lary who had been dreaming up cool technology in the Advanced Technology Group at Seagate, recently sold to Xiotech, because I figured that was what the company was on about.  Unfortunately, the former Seagaters were not in attendance.  Steve had the Indian flu and talked to us via telephone, probably from his bathroom.  I can’t say what about because of NDAs, but I can say that it is a very interesting development in storage and if played right could propel Xiotech to the front pages of the storage trades in short order.

What I did discover was something equally cool that isn’t under non-disclosure but no one has talked about in the trades that I am aware of.  Let’s jump back a second.

A month or so ago, I got a bunch of responses to my query to storage management vendors about the state of SRM.  Seemed like just about everyone had the same view of SMI-S — nice open approach to building a management interface on storage gear and all that, but not implemented widely or evenly by storage vendors who have little incentive to commoditize their wares with a common management scheme.  In short, SMI had succeeded in doing what it set out to do, but it wasn’t being used.  One fellow had remarked that going forward, he expected the attention that companies were paying to SOA and web services to drive an implementable solution for unified storage management as well.

Now, back to Minneapolis.  What I saw at Xiotech (and can talk about at this time) was their ICON Manager — the management utility that you use to carve up and allocate their array capacity to your apps.  This is a really good piece of code, folks.  It lets you do, in a linear and intuitive way, what has always struck me as an inordinately convoluted thing in Fibre Channel kits:  allocate storage to apps. 

So its a good management GUI.  So what?

ICON Manager sits on a web services platform that readily accepts value add integration from third parties, like Tek Tools Storage Profiler for trending and reporting.  This nifty implementation of web services outshines anything I have seen from SMI implementations and provides the capability for a storage vendor, in this case Xiotech, to enable its wares for feature/function augmentation via third party software and common management to boot.  While EMC, et al, are struggling to get some of the wind back in their sails lost to the knee knippers and anklebiters — adding thin provisioning, for example to their arrays this year — no one has paid much attention to this all important innovation in management by Xiotech. 

With the economy going to hell in a handbasket, management is set to become the darling of Corporate America as a centerpiece of CAPEX/OPEX containment strategy.  Xiotech is ahead of the pack.

Note that I have no vested interest in Xiotech.  They covered a plane trip up to Minneapolis and an overnight at a Marriott Courtyard and dinner at a chain restaurant.  I have not been paid to make any sort of endorsement.  But, I am very psyched about the web services based management utility I saw there.  It might well become the centerpiece of a manageable storage infrastructure if they decided to productize just the web services and management utility.

Worth a closer look.