Many IT pros believe that Linux and open-source software (OSS) is less costly to manage and maintain than Microsoft operating systems and applications. Naturally, Microsoft says that's a false impression.
So, which OS really has lower TCO? There are more than two sides to this question. For example, Framingham, Mass.-based International Data Corp., at Microsoft's behest, completed a study that, with qualifications, gives Windows the edge over Linux. In 2002, SearchWindowsManageability.com offered a preliminary report on that study and received about 100 letters from IT pros who either agreed or disagreed with IDC's findings.
In 2003, the debate is still raging, as evidenced in the "Open-source vs. Windows TCO" webcast, which features guest speakers CJ Rayhill, COO and executive vice president for O'Reilly & Associates, and Peter Houston, senior director of Microsoft's Windows server strategies group.
Facing more questions than they could answer in one hour, the speakers continued the debate after the microphones were turned off. In this Q&A, they discuss the TCO question. In part two, they cover the competition between the Linux/open-source camps and Microsoft. Next week -- in parts three and four -- they debate the challenges IT pros face in using or choosing between Linux/open-source and Windows and they analyze the successes and failures of the open-source movement.
SWM: Why is TCO important? Isn't the up-front cost of software an equally important factor in determining an IT solution?
Rayhill: The initial cost of software (and hardware) is far less important than the ongoing costs of staff, outside support, licenses and upgrades.
Houston: I agree that up-front costs are important, too, but since operating system deployments -- especially server operating systems -- have a very long life cycle, it's important to consider TCO as well. An analogy might be if a friend gave you a free car, and you found that it needed constant maintenance. It might end up being a better deal to have actually paid for a better car in the beginning.
SWM: Isn't TCO subjective? After all, every IT environment is unique.
Houston: We understand that everyone's environments will be different. We commissioned the IDC study to examine the elements of cost more deeply (and we hope that customers will read this study, if only to see what areas of costs they might need to consider more thoroughly in their environment) and to encourage greater industry debate on this topic.
Rayhill: TCO is certainly case-specific, but it's not subjective. It should be objective for your case, given the variables you have to consider. Which do you value more: the ability to customize an application, or a vendor-integrated suite? Your answer to that and other such questions weigh into the TCO equation for your business.
SWM: In a cost-conscious IT environment, what does the landscape look like for services-centric solutions/providers?
Rayhill: There are costs for services in either MS or OSS solutions. You must always budget for services and support, whether it's internal or external. Be careful when evaluating TCO to plug in the variable costs of services.
Houston: I caution customers to be careful about approaches that require a lot of services in order to deliver the final solution (and to maintain the solution). Such solutions might mean more revenue in the beginning to service providers -- and they might be tempted to propose solutions to their customers that require more services -- but, in the long run, I believe that customers recognize where they are getting the greatest value for their money.
SWM: What is the argument for homogeneity versus heterogeneity? Does heterogeneity impact TCO?
Houston: Heterogeneity is a way of life for most large organizations, and we recognize that here at Microsoft. But I would also say that heterogeneity doesn't necessarily need to be a goal -- and that greater heterogeneity generally translates to higher TCO. That's why we have focused our efforts on a more homogeneous approach, while simultaneously making sure we offer great interoperability technologies and build on open standards as much as possible.
Rayhill: You don't make business decisions based on homogeneity versus heterogeneity; you make them based on the needs of the business. Both homogenous and heterogeneous systems have advantages and limitations.
SWM: As Linux and OSS solutions continue to grow and improve, won't many of the issues -- such as management and support costs -- diminish, giving those solutions a TCO advantage?
Rayhill: Once again, TCO advantage is a case-by-case decision. The needs of the business should drive your evaluation of TCO.
Houston: This assumes two things:
(1.) that Microsoft will stand still; and
(2.) the kinds of advances that deliver improvements in manageability and supportability are being worked on in the open-source community.
Regarding #1, we are making continual advances in the areas of management and support, and Windows .NET Server 2003 [now renamed Windows Server 2003] will show considerable progress.
Regarding #2, I see most of the advances in Linux management coming from third-party vendors of management solutions. That leads me to think that managing Linux in the future might actually raise the TCO numbers because these third-party products tend to have high price tags.
Page 1 2
FOR MORE INFORMATION