|
Early this year, the Forrester research report, North America’s 2006 Enterprise IT Spending Outlook, cited 77% of firms as having rated the improvement of IT efficiency as their most important operational priority in 2006. Making matters worse, the Forrester report came on the heels of findings by Gartner Research in “U.S. IT Spending and Staffing Survey, 2005” that pegged 75% of IT budgets as going to maintain the status quo. Dig a little deeper and you’ll find that nearly 25% of the money allocated to maintenance goes into software support. If you dare to add in help desk costs, that number starts closing in on 40%. Is it any wonder that IT is often thought of as an impediment, rather than as an enabler to business efficiency and growth?
At the heart of this problem has been our basic notion of a computer application, which just so happens to be the cornerstone of the IT value proposition. The traditional view of a monolithic computer application is that it supports the manipulation of data for a number of interrelated business tasks that make up a business process or, at the very least, a significant portion of a business process. What’s more, the rules that govern the manipulation of data are a representation of existing business processes, regulations, and organizational structures. As a result, traditional programs embed complex interdependencies that represent the business processes in place at the time the application is put into service.
The crux of the problem for today’s CIO is that traditional notion of an application: It’s completely at odds with the notion of a modern On Demand Business. The backbone of an On Demand Business is the ability to quickly sense and respond to business-environment changes, which can range from supplier problems, to outsourcing issues, to market trends. To enable such a sense and respond business model, business processes long ago evolved from distinct silos into integrated end-to-end processes that are not tightly bound to a single organization.
The essence of a modern business process is that it starts with suppliers and ends with customers, with a great deal of interaction across organizational boundaries. Emerging business models feature both increased internal specialization and increased dependence on partners in extended value networks. In such a business model, processes are seen as dynamic rather than static activities.
For an application to meet the needs of an On Demand Business, it must exhibit two key characteristics: It must be flexible and easily modified, and it must be easily integrated with other business functions. IT must be able to quickly and easily assemble individual functional components into the precise application needed by the business at any particular time. As a result, monolithic applications give way to composite applications, which are woven from atomic business functions in what is dubbed a service oriented architecture (SOA).
A successful implementation of an SOA infrastructure requires the development of business functions with adaptable connections and well-defined, standards-based interfaces. Implemented correctly, an SOA infrastructure will encourage extensive reuse of existing services and assets and exhibit two critical value proposition characteristics: the loose coupling of existing services and rapid application deployment as new services are quickly put together from existing services with minimal new core development.
| What’s more, many experts in organizational dynamics would argue that SOA engenders greater organizational commitment through its focus on business-level activities and interactions rather than technical tasks. These experts also argue that this greater commitment helps to solidify an organizational bond between business and IT, thereby delivering greater business value. Nonetheless, the fundamental assumption that drives these benefits is that an IT organization is capable of executing an SOA strategy. That’s an extraordinary leap of faith given the results of a 2005 survey of IBM Tivoli customers, which revealed that 85% of IT service issues in day-to-day operations are caused by changes instituted by the IT staff. |
|
For a CIO, the greater complexity of SOA raises a number of very serious issues. Solving these issues is precisely where SOA intersects with IT Service Management (ITSM). The idea behind ITSM is to do things consistently, in an automatic way, and correctly the first time.
|
For the CIO, an SOA infrastructure, which features multiple services deployed on a common infrastructure, raises, rather than lowers, the fundamental complexity of the IT environment. Since new services are put together using existing services, there is an ever-present danger that deploying a new service could break a dependency of an existing service. While the possibility of breaking existing services is bad, the possibility of introducing quality of service problems, which are far more difficult to trace and debug is worse. For a CIO, the greater complexity of SOA raises a number of very serious issues:
- How can a better job be done by moving changes into the IT environment?
- Which services will be impacted by any given change?
- What can be done to ensure that IT resources are aligned with the services that produce the highest business impact?
- How can the business impact of a change in services be assessed before the change is made?
- What if a service for internal users must be reused for external customers?
Solving these issues is precisely where SOA intersects with IT Service Management (ITSM). The basis of ITSM is the development of an IT environment with the same kinds of quality controls that have evolved throughout the enterprise through such process improvement initiatives as Six Sigma. The idea behind ITSM is to do things consistently, in an automatic way, and correctly the first time.
To this end, IBM’s ITSM offerings have three dimensions:
Among the key process managers is The Tivoli Availability Management Process Manager, which is used to determine the business impact of incidents and deal with service level management, availability management, and incident management. At the operations level, IBM Tivoli Composite Application Manager for SOA provides proactive, real-time problem resolution. This is done through a portal-based end-to-end view of services, transactions, and associated resources across platforms and subsystems. IT lifecycle costs across operations, support, and development are reduced by directly linking operations and development.
Linking IT process with IT operations, the IBM IT Service Management Platform serves as an integration point for IT Service Management processes and management data. In particular it provides complete visibility into application complexity by automatically creating and maintaining application infrastructure maps. The comprehensive application maps include complete run-time dependencies, deep configuration values, and accurate change history.
In upcoming issues of Tivoli Beat we will be examining various aspects of ITSM and how it enables an efficient, cost-effective SOA environment. Included in these investigations will be a look at security as it relates to the role of the user. We will also be taking an in depth look at support for managing software releases.
|
