Environmental Engineering Reference
In-Depth Information
of a substantial amount of computing functions to relieve users of low level
management activities, allowing them to emphasize on the higher level con-
cerns of the pursuit of happiness, in general, or the activity of the moment,
such as playing in a soccer match, cooking a meal, or engaging in a spirited
scientific debate.
The need and justification for Autonomic Systems arise from the ever
increasing complexity of modern systems. A not uncommon complaint about
the information technology (IT) industry identifies its inordinate emphasis on
improving hardware performance with insucient attention to the burgeoning
of software features that always seem to require every possible bit of additional
hardware power, neglecting other vital criteria. This has created a trillion
dollar industry with consumers at the mercy of the hardware-software upgrade
cycle. The consequence is a mass of complexity within “systems of systems,”
resulting in an increasing financial burden per computer (often measured as
the TCO: total cost of ownership).
In addition to the TCO implications, complexity poses a hinderance
to achieving dependability [
156
]. Dependability, a desirable property of all
computer-based systems, includes such attributes as reliability, availability,
safety, security, survivability, and maintainability [
8
]. Dependability was iden-
tified by both US and UK Computer Science Grand Research Challenges:
“Build systems you can count on,” “Conquer system complexity,” and “De-
pendable systems (build and evolution)” [
60
]. The autonomic initiatives offer
a means to achieve dependability while coping with complexity [
156
].
8.1.1 What are Autonomic Systems?
An initial reaction to the Autonomic Initiative was “is there anything new?,”
and to some extent this question can be justified as artificial intelligence (AI)
and fault tolerant computing (FTC), among other research disciplines, have
been researching many of the envisaged issues within the field of autonomic
computing for many years. For instance, the desire for automation and effec-
tive, robust systems is not new. In fact, this may be considered an aspect of
best-practice systems and software engineering. Similarly, the desires for sys-
tems self-awareness, awareness of the external environment, and the ability to
adapt are also not new, being major goals of several fields within AI research.
What is new is AC's holistic aim of bringing all the relevant areas together
to create a change in the industry's direction: selfware, instead of the hardware
and software feature-upgrade cycle of the past, which created the complexity
and TCO quagmire. IBM, upon launching the call to the industry, voiced the
state of the industry's concerns as complexity and TCO. They presented the
solution to be autonomic computing, expressed as comprising the following
eight elements [
63
]:
•
Possess system identity: detailed knowledge of components
•
Self configure and re-configure: adaptive algorithms
