LAS VEGAS -- If you want to win the hearts and minds of enterprise IT managers, it helps to have a solid management story. Microsoft needed a better plan than the one it was pitching to its customers. So last year, the company hired a management guru who re-examined the company's products across the board. Microsoft also shifted product plans and people within its organization.
With some fanfare, Microsoft has rolled back the curtains on a new vision, which includes plans for an autonomic computing initiative and the integration of its three main manageability platforms. Don't expect big changes anytime soon, but the effort does show that Microsoft the company is certainly taking Windows management more seriously than before. David Hamilton, director of product management for the Windows management business group at Microsoft, discussed some of the changes this week at the Microsoft Management Summit.
|David Hamilton, Microsoft|
How many years before the System Center vision is truly realized?
David Hamilton: The Dynamic System Initiative is about approaching the crisis in the data center. Management is a big part of it. The initiative, by its nature, is a long-term commitment, like the Trustworthy Computing initiative. It's a multi-year, multi-release effort that doesn't end in any particular time. It's a broad philosophical commitment to make our applications easier to manage, easier to scale, more reliable.
The manageability components will happen in a number of different places. We are doing something with Visual Studio .NET to build things like the system definition model into the environment. Secondly, we are doing some things from the operating system to build the management infrastructure, and then with the System Center to build the management products.
It will start with SMS 2003 and MOM 2004. It gets further in Longhorn, and further still in Blackcomb.
At what point this year was this message developed? At last year's management summit, you discussed a completely different strategy.
Hamilton: For a long time, we've been trying to bring diverse pieces together. We've always had a Windows-specific focus. We have worked with different Windows teams. As we've done that, we've realized more that there is a better story on which to execute.
This year, there are two things that have caused us to refocus. First, the economic conditions are such that management has become a broader need for customers. And second is the arrival of Kirill [Tatarinov, corporate vice president of enterprise management]. We brought on a VP with plenty of experience in management, and he had a catalyzing effect.
How will you get senior managers to buy into this model? It sounds like it may be a different employee that would approve this technology, as opposed to a systems administrator that would work on SMS or MOM.
Hamilton: Today, management is an afterthought for most applications, and that's been the case for dozens of years. You've got to find a way to tie together management and the application. I think senior managers understand this. They think there is a real mismatch between application developers in their organization and their operation staff. They see the problems with the help desk, problems with application deployment and monitoring. And they want to reduce the costs associated with those problems. It's not a difficult story to tell, and I think customers want us to help them.
The new vision sounds like it will involve a lot of upgrades and added costs. Won't this hamper deployment?
Hamilton: Kirill refers to this as a revolution, but I think it has to be an evolutionary process, and that's why we are talking about this in stages. What we've done with SMS 2003 today helps. MOM 2004 fills in some key holes. And what we do with the System Center suite creates a bridging point that helps us move to this System Center built on the system definition model.
How exactly do you plan to merge SMS and MOM?
Hamilton: We will start with a suite that provides several levels of integration. First, the user interface is integrated. You will be able to display things in a common place and drive SMS actions off of MOM.
There will be integration at the data level. We will have a data warehouse that we introduce for the first time with MOM 2004. That data warehouse not only holds the MOM data, it will also hold the SMS data and allow you to draw some analysis and common reports across the two.
There will also be some commonality in terms of the installation. The suite by definition is a combination of the two products loosely coupled. We think there is some value to bringing them together at an early point. Over time, we will move that piece on top of the system definition. We will sell [MOM and SMS] separately as well at that point.
And Application Center goes away?
Hamilton: We will continue to sell and support Application Center until Blackcomb. But, gradually, the technology will move into the System Center product and into the operating system. Application Center focuses on three areas: load balancing, deployment and monitoring. We think those capabilities are more relevant in other places.
Many customers like simple point solutions for management. What is the argument for having a System Center versus having an element manager?
Hamilton: We are trying to make it easier for different element managers. What we see as the problem, traditionally, has been in broad framework plays that take a lowest common denominator and never really manage any other element well. We say we are going to be a great element manager of Windows and, in doing so, we will create good hooks for other element managers to snap in.
When in 2004 will the new MOM ship?
Hamilton: Summer is the best we can estimate, based on beta cycles and customer feedback.
What exactly is Microsoft's Indy project, and how does it relate to System Center?
Hamilton: It's a research effort for capacity planning. We have three research centers worldwide which are constantly churning out projects. One thing that came out was a capacity management project that went to MSN for a year to do some specific modeling around MSN environments to see if it worked. It did, so we are rolling it into the management division, and you will see it in System Center. The technology looks at the demands an application makes on a component and lets you dynamically work out how to change that assignment.
FOR MORE INFORMATION: