Finding the Web services 'sweet spot'

When I tell customers that my company does Web services management, the question I often hear is, "So, what do you mean by Web services management?"

It's no wonder that there's so much confusion on this issue, because the term "management" has been used to mean many different things. Two examples: business process management, the active coordination and execution of business processes; and systems management, the passive monitoring of performance and IT infrastructure). These are two very different meanings of the word "management"--and two very different markets.

Management in the context of Web services has gained a lot of visibility, as the big companies each try to claim ownership of the space.

		Get Up to Speed on... Web services Get the latest headlines and company-specific news in our expanded GUTS section.

Hewlett-Packard's OpenView division acquired Talking Blocks. Computer Associates acquired Adjoin. BEA Systems announced that its future WebLogic 9.0 release would focus on management. Are application platforms like BEA WebLogic and systems management tools like HP OpenView in the same market? Hardly. But both camps are trying to stake a claim on the space. Further confusing the issue, most pure-play Web services management solutions do a bit of both.

So who's right? Does Web services management belong in the domain of the application platforms? Does it belong in the domain of the systems management platforms? Is it truly a new market? Those in the industry state with absolute certainty that they know the answer. Unfortunately, their answers vary. As you would expect, it's not as clear-cut as any of the industry folks might have you believe.

In a world where the way people design, build and install applications remains constant, there is a clear separation between the roles of application servers and systems management tools: Application platforms run the applications; systems management tools make sure that the applications are running well. Application platforms are used by developers; systems management tools are used by operators. The two camps rarely cross paths. In a status quo world, what then happens to pure-play Web services management companies? In this world, consolidation is inevitable.

The open question is whether enterprise information technology is on the cusp of a fundamental change.

The open question is whether enterprise information technology is on the cusp of a fundamental change.

Web services alone will not cause a fundamental change. Web services can shrink costs and time to market, but those are just evolutionary uses of a new technology; not revolutionary.

The fundamental change that's coming in enterprise IT is the move to service-oriented architectures, which, in theory, enable IT organizations to avoid unnecessary duplication by sharing software services--such as order processing--across projects and teams.

Web services are the vehicle to make service-oriented architectures available to the majority of IT organizations. Through reuse and consolidation of duplicative services, a properly implemented service-oriented architecture results in significant improvements in both IT cost and flexibility. This reuse and consolidation changes the rules of the game in one important way: A software project is no longer self-contained. One IT project depends on others, which are maintained on different schedules by different organizations.

These cross-project dependencies result in new requirements that previously did not exist. Maintaining the overall performance and availability of the service network requires central visibility not only into the operation of all of the services, regardless of what platform they are built on, but also into active internal manipulation of network routing, quality of service and, in some cases, content of messages sent across projects.

Heterogeneous central visibility is foreign to application platforms but typical of systems management products.

Coordinating in-network control across heterogeneous projects requires a careful blend of functionality drawn from both systems management and application platforms.

On the other hand, active in-network control is something foreign to systems management products but typical of application platforms. Coordinating in-network control across heterogeneous projects requires a careful blend of functionality drawn from both systems management and application platforms. This is the "sweet spot" of Web services management.

This blending of functionality is what is driving both application platform companies and systems management companies to think that they "own" Web services management--even though, for the most part, the vendors are not building Web services management. As with anyone coming from a certain area of experience, their belief is that the problem entirely fits within their domain. To a hammer, every problem is a nail.

Systems management providers are building in passive monitoring of Web services, while application platform makers are building in functionality, such as routing--and both sides think that this gives them a complete solution for Web services management.

As the big companies become more aware that Web services management is a critical capability for maintaining a service-oriented architecture and that critical components of it are outside their experience base, they will scramble to bring the technology in-house, since only the largest companies have significant expertise in both areas--usually in product teams that rarely communicate with each other.

Regardless of whether a fundamental change is on the way for enterprise IT, the second- and third-tier Web services management players are likely to be shaken out of the market or acquired. The opportunity for the first-tier companies, however, is to reach escape velocity and become successful independent companies--demonstrating that Web services management isn't "owned" by any traditional market segment.