The concept of Web services has seen more than its fair share of media coverage over the last year. And so has the concept of service-oriented architectures, which is the use of Web services to define a model of loose coupling between applications. But the industry buzz regarding this latest "Next Big Thing" is surprisingly devoid of one important aspect that will mean the difference between the success and failure of Web services or service-oriented architectures - management.

Simply put, without the proper management of Web services deployments, enterprises will not see the business agility or improvements in developer productivity that they had hoped for. Instead, it is likely that they will find themselves in a chaotic mess with reduced qualities of service and higher costs to bring control back into their application environments. Such an outcome could seal the fate of Web services before they can deliver on some of their very real and achievable promises.

In traditional application development, management was an afterthought - the process went from develop to deploy, and only later did we think of managing what we deployed. As we started to deploy distributed applications the fallacy of that approach made us realize that management and manageability need to be considered much sooner. Web services, with their highly distributed, loosely coupled, plug-and-play, and oftentimes asynchronous nature demand that the process be develop, manage and deploy

Think of it in terms of airplanes and air traffic control. Sure, a pilot can fly a plane from point A to point B. But if you don't have air traffic control, you better hope no other planes are in the air. Even if there are no other planes in the air, the pilot still needs air traffic control to inform him of changing ground conditions. The pilot focuses on flying the plane, and air traffic control focuses on providing visibility and enforcing discipline consistently across the planes. With Web services, visibility and control are essential and are provided by a management platform that is to Web services what air traffic control is to planes. Like air traffic controllers, IT managers must have the tools in place that allow them to keep a pulse on a constantly changing environment and to take corrective action when necessary. And they need to take this corrective action without changing the Web service itself, just like the air traffic controller does not need to change any parts in the planes that are in the air.

There is another aspect of Web services and Web service management that is important: we need to think in terms of the life cycle of Web services - from development, to integration, to deployment, to change management. And testing, testing, testing throughout the life cycle. While this was important for traditional application environments, it is critical for Web services for the simple reason that the line between development and deployment is starting to blur. We now hope to have more "in-the-field" upgrades of Web services, more distributed, less coordinated change, and we will.

The management platform should provide seamless coordination among integration, management, and business analytics - helping companies respond quickly and intelligently to constantly changing business conditions.

Integration: Reduce Costs and Gain Control. IT organizations looking to leverage Web services in business-critical integrations need to define and enforce policies for security, quality of service, logging, and change management. A good management platform should provide a central console that reaches across multiple services - which can cut development time and keep managers focused on higher-level business logic instead of low-level plumbing, dramatically reducing maintenance and upgrade costs.

Management: Achieve Operational Visibility. The end-to-end management platform should also include monitoring capability to deliver operational visibility into Web services and Web services integrations - enabling IT managers to keep a close eye on critical service metrics across the enterprise, such as performance levels, downtime, and security violations. By knowing what is happening at every level of the integration - thereby improving application integrity, slashing problem diagnosis time, and enhancing security - enterprises can meet and exceed service-level requirements and ultimately improve customer satisfaction.

Analysis: Monitor Business Activity. Since Web services interactions can, and typically do, expose higher semantic information in the XML payloads and in the WSDL descriptions, the management platform should allow this business information to be processed and monitored. Therefore, a component of a comprehensive management platform is a business activity monitoring solution to provide business users with continuous visibility into important business performance indicators, e.g. number of orders, value of those orders, and number of goods shipped. With the ability to continuously monitor and analyze business processes, IT managers can proactively detect business problems or opportunities and rapidly notify appropriate business operations staff.

Enterprises know that to succeed in today's budget-constrained and highly competitive business environment, they need to be responsive to customer, partner, employee, and supplier demands - sometimes called a real-time enterprise. Achieving a real-time business state is possible with today's Web services technologies, but it does have risks. Only with a comprehensive management platform in place can companies hope to achieve significant ROI and fully realize the promise of Web services.