In the May 3rd issue of InformationWeek, Bob Evans wrote an article “Oracle’s Phillips: Standardizing On Oracle Is IT Cure.” I am sure most IT companies won’t agree with it even though Oracle is now a full stack company after grabbing Sun MicroSystem not long ago. The big players probably want to claim the same for themselves, for example, standardizing on IBM is the IT Cure.
Digging further into the article, we can find some interesting arguments by Phillips:
Lost VMs or Containers? Too Many Consoles? Too Slow GUI? Time to learn how to "Google" and manage your VMware and clouds in a fast and secure HTML5 App.
What CIOs are struggling with right now is trying to find a way to get the opportunity and ability to manage the entire stack with a single management tool that’s predictive about that stack’s going to behave, how the change management around it is more prescriptive and planned, and where they really know how to upgrade and patch the entire stack.
All the dependencies between these layers – the middleware, database, storage, software, systems — they’re all related but unpredictable. And that’s the cycle they’re trying to get out of it — all that need to constantly provision and manage — it’s a huge cost, and it’s kinda boring and takes lots of people to do it, and it’s risky.
CIOs are ready for a new approach. The entertainment value or intellectual stimulation you get from tweaking every little thing up and down that stack is no the same as what it once was. They’re kind of bored with with doing that. And the expense of doing that is apparent now, and after going through the last two years of downturn, it kind of helped us in a way because people said, “I’ve gotta find a way to change what I’m doing. That is not working.”
You don’t need 18 different vendors and 2,000 configurations to have competition. You got to limit it some. And I think we’ve convinced people that makes sense, and beyond that, we think the whole industry’s just moving in that direction. And we can accelerate that by standardizing that entire stack and showing people how it’s done — people like that “iPad for the enterprise’ analogy.
I think all the above statements are perfect testimonial for another alternative cloud computing where “you got to limit it Zero.”
Although the normal customers won’t have to build in house infrastructure in cloud computing scenario, the service providers will take over the pain and gain. When a service provider is big enough, it will have strong interest to build its own stack with different components for best cost structure. Quite of these components will be open source projects with or without paid support.
For the cloud computing customers, they don’t need to worry about stack then. As the authors of the book “Are You Lights On?” pointed out, after you solve one problem, a new problem will arise. The new problem for the cloud customers is how to avoid service vendor lock-in. I will blog about the strategies later.