Dynamic evolution of business performance management solutions using declarative evolution policies

摘要

A policy-driven approach to Business Performance Management (BPM) evolution is to capture the mechanism of BPM solution evolutions. These policies are executed by the runtime infrastructure which transforms and interprets evolution policies and allows fine granularity controls on solution evolution. At transformation time, the relationship among the event processing progress and the runtime state is derived based on the observation model. Such relationship information allows verification of freshness of migrating data and parallelization of runtime state validation and event processing. Consistent and efficient evolution of BPM solutions transit while event processing is operational.