| Title: |
Autonomic Computing |
| Author: |
Richard Veryard |
| Publication Date: |
21 January 2003 |
| Report Type: |
Journal |
| Report Class: |
Best Practice |
| Abstract: |
The growth of software componentry and Web Services, and the continuing ambitions of eBusiness processes and services, lead inexorably to a step growth in the scale and complexity of enterprise systems and collaborations. Platform providers present a variety of responses to this challenge. In this article, we examine two apparently contrasting approaches - the virtualization approach found in Sun's control plane, as well as in HP's "OpenView" product; and the development of autonomic computing, which is at the core of IBM's strategy. |
| Backgrounder: |
Life doesn't get any easier. As fast as technology removes some of the difficulties and obstacles of business computing, the demands and ambitions of business increase to maintain the challenge for IT management. (Perhaps surprisingly, the challenge remains at a fairly constant level over time, irrespective of technical innovation).
Web services and related technologies are also used to facilitate utility and grid computing, and thus enable a further increase in scale and complexity.
System componentization and complexity generate a number of problems. Complex systems can be more difficult and expensive to install, to tune, to operate and to alter. Most enterprises already have far more complexity in their internal and external systems than they can manage comfortably and effectively. This situation compromises efficiency at many levels - as well as endangering reliability and security.
As size and complexity of IT applications increase, the size and complexity of system management increases by at least an equivalent proportion.
Some vendors deny this. They point to technological developments (including standards) that seem to alleviate or eliminate some aspects of complexity in some areas. And we must certainly give the IT industry credit for many local simplifications - from Web Service standards that simplify connectivity between diverse systems, to grid computing that simplifies the sharing of resources across large heterogeneous networks.
However simplifications at one level often merely expose difficulties at the next level. For example, syntactic standardization exposes the need for semantic standards.
Furthermore, as simplifications at one level remove certain obstacles to scale and complexity, systems are pushed into larger and more ambitious requirements until similar (or worse) obstacles are encountered somewhere else.
Meanwhile, advances in system management and other devices may naturally lead us to increase the levels of componentization of infrastructural elements. Component techniques may be essential in order to achieve the levels of reliability, security and adaptability - but these in turn may escalate the demands on system management.
|
| Report Size: |
11 Pages |
| Report Access Type: |
 | Silver/Gold (Premium) |
|
| Available for separate purchase |
Single copies of recent CBDI Journals may be purchased |
| Login |
|
Home