Self-Management and the Many Facets of ‘Non-self’

The difficulties in dealing with increasingly complex information systems that have to operatein dynamic operational environments calls for self-management properties or, more ingeneral, for the integration of “self-*” features (e.g., self-configuration, self-adaptation, selfhealing)in software and information systems.The common perspective of nearly all “self-*” approaches is that of considering humanbeings as “non-self” from the information system perspective (Figure 1-a). Indeed, one of thevery common goals of all approaches is to move humans out-of-the-loop, by makinginformation systems able to perform in an autonomous way all that kind of (costly and oftentoo complex to bear) human activity related to configuring and maintaining informationsystems so as to have them properly working under all conditions. However, beside thiscommon perspective, a number of diverse conceptions co-exist for what should be actuallyconsidered “self” and what, beside humans, “non-self”.