|
|||||||
|
|
![]() |
|
|
IBM Introduces Autonomic Computing Blueprint and Cure for 'Spaghetti Code' by Alex Woodie At IBM's developerWorks conference in New Orleans last week, Big Blue talked about two new initiatives that seek to bring legacy environments into the present and to make self-healing autonomic computing technology a reality. Big Blue detailed its blueprint for autonomic computing, including tools based on open standards that it will encourage its partners to use. In a separate announcement, new software and services were introduced that modernize "spaghetti code" running on customers' proprietary mainframe and midrange servers. IBM's blueprint for autonomic computing defines a set of technical guidelines that will allow IBM and other vendors to ensure that different components of a self-managing infrastructure work together. The blueprint begins the process of developing a common approach and terminology for architecting autonomic computing systems, and introduces a set of consistency mechanisms and control loops that developers can use to ensure the integrated roll-out of self-managing capabilities across the enterprise. IBM announced the blueprint a week-and-a-half ago, and revealed additional details about it at the show last week. Along with the blueprint, IBM has introduced four new tools--all based on open standards and technologies--to promote interoperability of vendors' autonomic software. The tools include the Log & Trace Tool for Problem Determination, the ABLE (Agent Building and Learning Environment) Rules Engine for Complex Analysis, the Monitoring Engine providing Autonomic Monitoring capability, and Business Workload Management for Heterogeneous Environments. It was unclear at press time what applicability these tools would have for Windows and Linux platforms. While the software and services related to its "spaghetti code" announcement don't directly affect Windows and Linux platforms, Windows and Linux servers have a secondary play. "Spaghetti code," according to IBM, is code that was written 20 to 30 years ago--usually on a proprietary midrange or mainframe machine--and has been modified so many times, by so many people, that it is difficult to untangle it to make further changes or improvements. But this legacy code is vital, as it underlies applications that run 30 billion transactions per day, or about 70 percent of the world's major business operations, according to IBM. Companies have spent well over $1 trillion in developing this code (that's a figure IBM had cited before the Y2K jump at the end of 1999). They are loathe to spend more money than they really have to, and this limits how they can adopt and adapt to new technologies as they are introduced, including middleware running on Windows and Linux platforms. Recognizing there is a problem is the first part of solving it. IBM is rolling out a collection of software tools and services to help companies identify their spaghetti code and create plans to modernize and upgrade the business applications that they run. IBM Business Consulting Services has developed a tool its consultants can use to estimate how much a company can save by engaging IBM in either one of two services offerings. Those services include an offering called Application Portfolio Management Services, in which IBM consultants evaluate all of an organization's applications and recommend which ones to keep, abandon, or change. The second services offering, Legacy Transformation Services, seeks to renovate, integrate, migrate, and Web-enable these legacy applications. IBM is also announcing new tools to help developers bridge the gap between older OS/400 and mainframe applications and new applications running on WebSphere, which is generally deployed on Windows and Unix servers these days. IBM says the tools, which support major Web services standards, can be used to help update older IBM mainframe CICS applications and OS/400 green-screen applications to run on the Web.
|
Editor
Contact the Editors |
| Copyright © 1996-2008 Guild Companies, Inc. All Rights Reserved. |