VP Kirill Tatarinov talks about Microsoft’s strategy with the System Center product family
Industry pundits roll their eyes when Bill Gates says that IBM, not Linux (or Google), is Microsoft's biggest competitive threat. Those pundits assume Bill is being disingenuous or diversionary with such statements. But I think he's not. Open source in itself doesn't endanger Windows' march toward being the predominant business-computing platform. Rather, the danger lies in the potential availability of a standardized stack of integrated and easy-to-manage open source core applications emulating the Windows Server System (WSS) family—and IBM is the purveyor of integrated open source solutions.
At least since the late 1980s and the AS/400 platform, IBM has understood the power of a unified platform that seamlessly integrates essential applications (e.g., database, networking, and messaging) and reliably serves end users' needs while—most significantly—simplifying development and operations. Today, IBM's business thrives on deploying its skilled and knowledgeable services division to bring such unification and integration to the diversity of Linux offerings. By spreading open source skills to IT organizations, IBM is an ever-present reminder of the need to keep a competitive edge by further integrating theWSS application stack and continuously simplifying and unifying the management experience throughout the stack.
Assimilate IT Skills and Knowledge
Microsoft is addressing this threat by incorporating IT knowledge and skills into its products while simplifying management of a better-integrated Windows stack. A company wide strategy focuses on the complete life cycle of application development, IT operations, and end-user productivity. The elements of this strategy are the .NET Framework, the Dynamic Systems Initiative (DSI), Trustworthy Computing, and the Digital Workstyle for the New World of Work.
The piece of this strategy that directly speaks to capturing Windows platform skills and making them widely available is DSI, with its promise to "radically simplify IT" by reducing complexity and total cost of ownership (TCO), thus evolving IT from cost center to strategic asset. The key to DSI is expressing systems and IT knowledge and policy as models that connect the requirements of developers, IT pros, and end users. The goal of DSI is to produce self-managing systems by enabling developers to write "operationally aware" applications with Visual Studio (VS) on the "operationally aware" platform of Windows Server (and Virtual Server). Microsoft's System Center management products are the tools that capture IT knowledge and skill and make IT's life easier.
In a recent briefing on DSI, Eric Berg, a director of product management in Microsoft's Windows Enterprise Management Division (WEMD), told me, "The key with DSI is thinking about all of the information and knowledge about the application that resides in the heads of customers. Take a typical distributed application: Our customers have a lot of knowledge about the complexity of that application and the interdependencies of the application's components. Administrators are deeply steeped in the black magic of running their applications and maintaining them. So how do we take the knowledge, capture it in the software, and let the software automate it? We're capturing that knowledge in software models. A consistent theme is to make sure that everything we're doing includes not only operations or not only development, but connects them throughout that life cycle."
With this background on the DSI initiative, I talked with Kirill Tatarinov, Microsoft corporate vice president of WEMD. We discussed the System Center management solutions that Microsoft is developing according to the overall company strategy.
Connecting IT with Business
KF: The DSI strategy is bringing some major changes to Micro-soft's management products. What drove those changes?
Tatarinov: Management was historically thought of as occurring after the fact. People would build systems that were unmanageable and then try to manage them after the fact, approaching them as black boxes, and it just didn't work. You can't manage the unmanageable.
Organizations increasingly realize that IT needs to stay very closely connected with business, and IT agility comes to the forefront. At the same time, the drive to reduce cost continues to be prevalent, as it has been since the late 1990s. It's a conflict for organizations to reduce cost on the one hand and increase agility on the other hand. To do that, you have to really connect your processes, products, and people.
KF: What are you doing differently now to accomplish that?
Tatarinov: Starting with the process, we're helping IT managers understand how they should drive incremental improvement and how they should drive standardization and rationalization of their organizations to the point where they both reduce cost and achieve agility through implementing the process, implementing the product, and getting their people trained in both products and process.
Fundamentally, it's about connecting IT with business, about being able to dynamically respond to business needs, to quickly implement new processes and new business applications, and really to keep IT under control.
System Center Products
KF: How does your strategy for the System Center products and other management tools support this overall initiative?
Tatarinov: We have a very strong product lineup coming out through 2006 and the first half of 2007. This will be the time when we ship the next version of Microsoft Operations Manager (MOM), which we currently call MOM Version 3. We will ship the next version of Systems Management Server (SMS) with new features such as configuration management and a completely revamped UI. We will roll out model-based management, where you will be able to manage services end-to end in MOM and model services end-to-end with SMS. People increasingly want to manage their services, as opposed to individual components. Customers want to see something like their messaging service managed end-to-end and presented as one entity, as opposed to Exchange here, Active Directory (AD) here, DHCP there. That's the direction we're taking with our management products. You'll see a lot of that in MOM Version 3. You'll see concepts of that in SMS Version 4, rising out above the infrastructure components to infrastructure services. \[See "System Center Product Roadmap" for details about upcoming products and new releases of SMS and MOM.\]
KF: SMS and MOM are tools for larger enterprises. What about management in the small-to-midsized business (SMB) space?
Tatarinov: As far as management is concerned, small and medium business has been grossly underserved. In studying this market, we saw a tremendous number of people trying to roll their own applications or downloading open source applications to build solutions themselves. We're addressing this lack of tools.
Our first entry into this market was MOM 2005 Workgroup Edition, a simplified version of MOM that is priced for that market and simplified in terms of deployment and customization. We learned a lot. Based on customer feedback, next year we're introducing a product called System Center Essentials, which combines all the capabilities IT generalists need for managing their organization in terms of software update and distribution and patch management, as well as traditional MOM functions like event management, alerting, and monitoring.
KF: What specific features are in System Center Essentials as a result of the customer feedback you mentioned?
Tatarinov: Specific feedback was around breadth of functionality. MOM monitors operations. It doesn't do anything in Change and Configuration Management (CCM), software distribution, and patch management. With Windows Server Update Services (WSUS), we have good technology for that domain in the midmarket. WSUS has an easy-to-use interface designed for the mid-market, but it wasn't integrated with MOM. We are bringing the two together in System Center Essentials.
KF: System Center was originally announced as a product that would combine MOM and SMS, but last year you announced that System Center will brand Microsoft's family of management products instead. However, System Center Essentials sounds like something of a marriage of MOM and SMS capabilities. Is it?
Tatarinov: You could call System Center Essentials an evolution in that direction. We clearly see the emergence of a "versatilist" persona, as the Gartner Group calls it. This is the midmarket IT professional who manages broad IT infrastructure as opposed to just one discipline. This is the persona System Center Essentials addresses. In larger organizations, we still see specialists focused on operations management and CCM or data protection management, and we address those specialists with unique products designed for them. We will continue to evolve with the market.
Management Pain Points
KF: What do you consider the biggest management pain points for IT?
Tatarinov: I'd say deployment is the worst. Although Microsoft and others have deployment technology, deploying applications and updates, making sure people have the right software, making sure the software is compatible with other versions of other soft-ware—that whole area still requires a lot of work. Improving deployment still requires innovation and a lot of technology.
KF: I think one of the biggest pain points is errors and events. What are your plans there?
Tatarinov: Historically, neither Microsoft nor non-Microsoft products have done a particularly good job designing errors so that they're intelligible and actually tell you something. So in Windows Vista, we have made a major effort to revamp both event management and the error reporting system itself, but also the actual errors and events. Every group inside Microsoft has done a scrupulous job going through all of their event messages, building event manifests, and making sure that when an event gets triggered it actually gives the administrator or the end user some intelligible information about why it happened and what needs to be done. We're also making sure that this event manifest is connected with Tech Centers on TechNet so people can automatically connect back to our Web site and find a Knowledge Base article or whatnot to help them diagnose or troubleshoot. And of course, we take all event and alert information and use it in the construction of MOM management packs to make sure we can further automate event management for organizations that run MOM or System Center Essentials. So yes, it's indeed a big problem. We've invested heavily in solving this problem and I'm sure hoping this problem will go away.
KF: How does error and event management fit into WEMD's priorities?
Tatarinov: It is definitely a core part of the management business here. We provide all the infrastructure, we also drive and program-manage all efforts across the company to revamp events and make sure events are much better manifested and errors are intelligible—just like we drive similar efforts for creation of MOM management packs and desired state models. A variety of management efforts are done across the company but are led from the management group.
Designed for Operations
KF: DSI is a huge initiative for Microsoft, yet 84.5 percent of Windows IT Pro readers surveyed said they were not familiar with DSI. What should IT pros think when they hear "DSI"?
Tatarinov: Well, we obviously want them to feel great when they hear DSI. It makes their life easier, gives them control, finally puts them in charge of IT infrastructure, gives them visibility across the ecosystem and across the life cycle of the application.
KF: When I spoke with Eric Berg, he told me that Microsoft is setting the example with its own products. He said, "In Visual Studio, we're working to make applications more designed for operations. And we're taking the applications we deliver—Exchange, SQL, BizTalk—and making sure the models for operations are designed into them. We're both helping people who have to develop their own applications and then making sure that the applications we deliver are designed for operations. On the platform side, we're making a lot of investments in Windows itself, as well as the virtualization infrastructure, to make sure Windows is the right platform for these dynamic systems." Do you want to add anything to that?
Tatarinov: The most important, critical piece of DSI is the concept we call Designed for Operations, which means developers have to build things in a manageable way. We want to explain to people that this is really how we built Vista. Specific examples include manifesting events and making sure everybody builds health models for every single component and that those health models can be automatically moved into MOM management packs. This whole concept of building manageable applications and systems, including models, at design time and then being able to link those models to the operational side is essentially what DSI is all about.
Radically Simplify IT
Like IBM, Microsoft has always understood the value of a unified and integrated platform. That value was the idea behind the late BackOffice Server bundle (not to mention its descendent Small Business Server—SBS—as well as the upcoming Centro) and also was
the reason behind Microsoft's recent messaging about how the WSS products (e.g., Windows client and server, Exchange Server, SQL Server) are "better together." The argument is that you can build a Linux-Apache-MySQL-Perl/PHP/Python (LAMP) stack, but you don't get the optimization and integration that Microsoft's stack provides.
What Microsoft hasn't been able to do up to now (not least because of its siloed corporate structure) is to provide a consistent management UI and experience throughout its stack. It looks to me that Microsoft has realized that by integrating IT knowledge and skills with a consistent management UI, the company has a way to hold off open source advances into Microsoft's customer base—and that realization has finally provided sufficient motivation for Microsoft's various product teams to work together on an overall management strategy.