It's no news that I'm a contrarian. I don't do it intentionally; I just think most people are lemmings, so by the time everyone comes around to my point of view, I figure it must be time to change it.
Sun Microsystem Inc.'s storage has been a joke for a long time. Now it seems most everyone thinks that way, which is why I must now tell you that the odds are that you are completely wrong.
Sun has had lots of problems in storage. It didn't make good stuff. It didn't buy good companies. Otherwise things were fine. Scott McNealy said, "Storage is strategic," and then went to play golf. Ed Zander said, "Storage is our strategic growth area," and then went to play with Motorola.
A few years ago Sun started to put their money where their mouth was. It put Mark Canepa in charge. Canepa was a successful Sun bigwig from the server business. Canepa inherited a bunch of Sun old-schoolers, and for a while it was business as usual. They made bad products (engineering tended to be the ones to decide what was built never a good idea) and marketed them poorly (engineers tended to run marketing never a good idea).
Canepa slowly but surely has replaced a ton of old-schoolers with some smart new blood. They have marketing now. (I can hear you laughing, but I'm serious!) James Whitemore is not only smart and a realist, even better, he's a sales guy. How novel, putting someone who's actually spent time trying to sell stuff in charge of what products we need! Doug Wood from Pirus runs architecture. An outsider? About time, if you ask me.
Anyhow, Sun couldn't build an EMC killer, so it OEM'ed one from HDS. They couldn't build a low-end box cheap enough, so they OEM'ed stuff from Dot Hill and Engenio. They know now that the strategic value is not the pieces at the end; it's the smarts in the middle.
While IBM and EMC and Veritas and Cisco (etc., etc.) get all the virtualization ink, it's really Sun that has the best grid storage messaging right now. The concept of the "data abstraction layer" where applications are abstracted from the physical devices and virtual services execute in the fabric is not new, but putting it into practice is. IBM has shipped more than 1,000 SAN volume controllers, which makes them the king right now, but Sun has even grander aspirations. In reality, outside of IBM and HP, there isn't anyone else who has all the necessary pieces to pull it off, those pieces being server/storage/networking technology, sales force and, most importantly, customer base.
Can they make it happen? I don't know. I'm encouraged by the progress the storage unit has made (actually, I'm amazed), but there are other things that can screw it up. Wall Street hates Scott McNealy, mostly because he hates Wall Street. That keeps the company from being able to capitalize itself effectively or to grow aggressively.
It takes a worried man...
And judging from some of his rambling blogs I've read recently, new COO Jonathan Schwartz, the pony-tailed boy wonder, could be clinically insane. The fact that the company would even let these blogs out from their COO concerns me greatly. I worry that smart people will leave in this kind of situation. I worry that the powers that be spent so much time bashing Microsoft for no good reason that they missed the bigger picture, i.e., say what you will but commoditization is not only real, but it's good for society. Wake up and smell the Linux.
I worry that the storage guys might actually build the best stuff ever, but the sales force is so screwed up that no one will ever know. I worry that even as I write this and give Sun accolades, they go and buy Procom, an awful, awful company. I worry that Sun with 11 billion smart engineers needed to buy an awful little NAS company, even when it owns NFS. But as they say, you can't turn a big ship quick, or some such.
So, call me crazy, but I think Sun will be one of the most interesting players in the always interesting business we call storage. It has a real opportunity to change the landscape, and let's face it, it has nowhere to go but up.
So, can they do it? Sure they can. Will they? No idea. What about HP? They are even more screwed up than Sun in many ways, but I think that both have the ingredients necessary to make killer comebacks.
At the end of the day, it begins and ends with talent. Not engineering talent (sorry engineers, I really don't mean to bash you as much as I am; perhaps I'm just jealous), but business talent. Egos need to move aside, and logical intelligent people properly incented and ruthlessly supported are what Sun and HP need.
EMC, Netapp, Oracle, Veritas, etc. have killer people, and that is why they are successful. Mediocre talent blames engineering for their shortcomings. Killer talent doesn't care what engineering does it goes out and finds the pieces it needs to get back into the game.
Steve Duplessie is senior analyst and founder of the Enterprise Strategy group. He can be reached at [email protected]
Find your next job with techworld jobs