Quality-based Approach to Urgent Workflows Scheduling

Urgent computing capabilities for early warning systems and decision support systems are vital in situations that require execution be completed before a specified deadline. The cost of missing the deadline in such situations can be unacceptable, while providing insufficient results can mean an ineffective solution that may come at a very high cost. In order to provide a solution that is appropriate under the current conditions (i.e. available volume of computational resources, workload, and time available), a new approach is required. In this paper, we present a schema and algorithm of regulating the volume of computations within an urgent workflow to deliver a solution that is as sufficient as possible given the current conditions and deadline. To achieve these goals, we develop an approach that modifies an urgent workflow by changing its structure and the parameters of its individual tasks. Such modifications are based on introducing a notion of quality and applying quality-based models to estimate the sufficiency of solutions generated by the resulting workflow structures. Finally, a special extension of the genetic algorithm that performs quality-based scheduling of urgent workflows is described along with an experimental study to demonstrate its efficacy.