Web Services Management Update

Organizations are proceeding faster with Web services implementations than expected. As a result, organizations will begin seeking Web services management (WSM) technology in the not-too-distant future. However, this is a fast-moving market with developments in standards, tools, adoption, and processes.

META Trend: During 2003-05, new infrastructure and application architectures (e.g., Web services, virtualization, utility computing) and budgetary restrictions will drive additional focus on capacity planning and meta-management efforts (e.g., status aggregation, business views), leading to more integrated meta-management tools (2007). Through 2006, continued cross-boundary management demands (e.g., organizational, informational, technical) will drive process, sourcing, and instrumentation changes.

A recent META Group study indicated 74% of respondents agreed that Web services are the future, and 64% either have implemented or are in the process of implementing some form of Web services strategy. As often happens, management of a technology becomes a need after mainstream deployment of the overall technology. Although many organizations are implementing or experimenting with Web services, a wholesale conversion of applications built entirely on Web services has not occurred. Rather, there have been opportunistic deployments for specific purposes (e.g., integrating existing applications). Therefore, mainstream demand for Web services management technology will not occur until late 2004. For the time being, WSM has established itself as a standalone management category. This market is clearly in flux with rapid change in standards, tool deployment options, and the lack of clarity in areas such as pricing. For organizations seeking to invest in WSM technology deployment options, the directions of standards and the future of the WSM category must be understood to ensure that the correct investment is made.

WSM is a confusing term. It is primarily used to refer to operational management technology (traditionally monitoring) but with expanded capability (e.g., security, load balancing, provisioning). On the development side, WSM is used to refer to the support of the developers (e.g., unit testing, coding). Such expanded definitions are clearly creating confusion among end users. We believe WSM technology is used to support the running of a Web service environment and eventually service-oriented architecture (SOA). WSM is and will remain broader than traditional management. However, this will cause some tool acquisition difficulty, as current Web services efforts are restricted to a small self-contained team with autonomous ability to select tools, even if they conflict with other organizational decisions (through 2006). Because Web services are more broadly deployed, different IT organizational silos will have a hand in key areas (e.g., security, operations, engineering), and each is accustomed to procuring its own tools and working independently. Gaining consensus among the various stakeholders on a single WSM solution will be difficult.

Tool acquisition difficulty is compounded by the overall direction toward an SOA for applications. When most applications are based on an SOA design (e.g., many discrete elements working together to perform a business service) and leverage Web services in some form (although this may not happen for 7-10 years, it will be an architectural direction for most organizations), a singular monolithic management solution is not feasible.

We believe the WSM market will cease to be standalone by YE05. In its place will be a service oriented management architecture (SOMA), on which many of the current WSM vendors will work to provide pieces of the puzzle. This SOMA has a chance at success because it will be based largely on standards, while past efforts were largely proprietary. There will be no single framework technology as the basis for a SOMA but rather a set of understood management technology sharing information (real time and historical) based on standards. We have referred to a specific instance of a SOMA as adaptive resource management. Therefore, users will not be forced to vendor lock-in. In addition, there will not be a single SOMA in an enterprise. Rather, different uses will enable different SOMAs to exist, possibly for dissimilar purposes (e.g., a SOMA for one division and another for monitoring) with standards-based interfaces to enable them to communicate.

Standards. Management (e.g., monitoring, configuration, change) typically does not do standards well. However, there are an amazing number of standards in the Web services domain that are either directly targeting management (Web Services Distributed Management [WSDM]) or have management implications (e.g., Web services provisioning). In addition, there are numerous bodies cooperating to make these standards compatible and complimentary, as opposed to competitive efforts of the past. Oasis is leading the way in the creation of Web services standards, with the core of the effort called WSDM. WSDM has several goals, staring with making Web services a manageable resource by leveraging Web services architecture and technology to manage distributed resources. Management covers a broad set of operational processes and defines how to interact with Web services to do things such as change management, configuration management, and how to discover and monitor Web services (e.g., health, availability, performance, usage). The goal is to perform all this management leveraging Web services technology itself. Oasis is also involved in other standards creation (e.g., Web services provisioning) that has management implications.

Besides Oasis, the Distributed Management Task Force (DMTF), Global Grid Forum (GGF), and worldwide Web Consortium (W3C) are contributing to the overall Web services management effort. The DMTF is contributing the Common Information Model (CIM) as a schema data-sharing foundation. The W3C is contributing the definition of state of an individual Web service (e.g., what is the correct term to describe a Web service that is active and ready to begin processing?). The GGF is involved in ensuring its cornerstone standard. The Open Grid Services Architecture (OGSA) has several concerns, including the following:

• Developing a foundation for virtualizing resources that can be consumed in a collaborative manner
• Using Web services to expose a standard interface and identify a consumable IT resource (e.g., networks, databases, storage, computational)
• Using standard interfaces to interact with those resources (e.g., how to move work between servers)

The working group is supported by many large organizations (e.g., IBM, HP, Dell, Computer Associates), and there is cooperation among the different bodies. Although these standards are complex (10-20 different standards, four plus standards bodies, and most of the work is in progress), we are encouraged by the cooperation and clear lines that seem to exist among the efforts. We are also encouraged by the depth the standards have taken. We are optimistic that real results will be achieved but do not expect any tangible action prior to YE05. Deviation by any key vendor (e.g., IBM, Microsoft, HP) or lack of progress during the next six months could hurt or kill the effort.

Study Findings: What will be key to the standards success is the inclusion of management concepts at the time of development. Although this has not been how management is traditionally handled, a recent META Group study on Web services reveals that organizations have begun to realize they must approach Web services management differently. The survey shows that 5% of respondents identified management as very or somewhat important, a high number for technology this immature. Even more encouraging, 51% identified that management capability is expected to be built into the Web services themselves, with security (79%) and performance monitoring (60%) as the two most important concerns.

The combination of changing user views and solid standards efforts will aid change in the WSM market during the next two years. Management will become more a part of the development process (even automatically generating management in the integrated development environment), standard data will be available, and standard administrative functions will be exposed. All this means that organizations currently seeking Web services management solutions must view those as either tactical purchases or ensure that the selected vendor illustrates a clear understanding of these long term issues and will evolve the current product set with clients (and not require additional major purchases).

Business Impact: Any savings from a Web services effort will be lost in the long run if the ongoing support of that effort is not accounted for in advance.

Bottom Line: Companies embarking on Web services projects must understand management implications to ensure logical investments in supporting management technology.

META Group originally published this article on 17 December 2003.