Introduction

Current-generation enterprise systems are complex IT ecosystems that serve mission-critical business functions. The set of applications that constitute an IT ecosystem are not simply a bunch of independent applications with application-level integration schemes and protocols; instead; they share complex interdependencies among them—they are usually integrated at a platform level instead of at an application level.

Today’s enterprises are technology-driven organizations—the technology enables the enterprise to function as a multi-function, multi-process, and multi-structured organization. Appropriate organization structure and related business processes are designed to serve a business function in the most effective fashion possible.

Enterprise architecture primarily refers to how to design composite, standards-compliant software applications and systems that fit into an overall enterprise engineering goals and objectives. Technology has matured immensely in the last few years. Most of the domain problems of yesterday are routinely solved as “implementation” problems today. More importantly, the emergence of standards in the way technology is conceived, created, deployed, and managed has enabled many product and services companies to offer integrated services to end-user enterprises, thus creating many business opportunities.

However, most knowledge about enterprise architecture is primarily limited to how to conceptualize, create, build, and deploy standards-compliant, loosely coupled applications that are integrated at a platform level. There is very little information or knowledge regarding the post-deployment management issues that arise when effectively running a complex enterprise IT system.

Many software engineering methodologies and disciplines address issues only until the point of verification and validation; they tend to remain silent about post-deployment management complexities. Most software engineering methodologies do not address the supportability aspects of the software. As a result, we do not have proper models to estimate the cost and effort of supporting complex application systems.

There are standards, such as ITIL, but these standards are mainly limited to infrastructure management and they fall short in the case of complex distributed systems management.

The following issues make the management of enterprise IT ecosystems even more complicated:

• There is an increased trend toward composite applications and SOA. This enables building highly configurable run time assemblies. The holy grail of SOA and composite fusion techniques is to push the configuration of the system from the design and implementation time to the post-deployment time. This increases the complexity of management of enterprise systems by at least an order of magnitude.
• There is also an increased trend toward outsourcing the management of the entire enterprise system. However, an effective outsourcing model requires a carefully designed methodology to define the SLAs capture the various management challenges, and to provide mechanisms to integrate the applications in an overall application/system management environment. There is not a unified methodology to manage measure and monitor the enterprise systems.

The goal of this set of articles is to fulfill this gap. In this series of articles, we describe a systematic approach to enterprise IT ecosystems management.

This article introduces SOA and composite techniques from the context of enterprise applications management. We argue that we must radically change our existing definitions of infrastructure—enterprise infrastructure must include all resources that are part of the ultimate service delivery, not just the physical hardware. We describe various issues that arise when creating a successful enterprise applications management practice.

Pages: 1 2 3 4 5 6 7 8 9 10