Evolutionary Inheritance in Workflow Scheduling Algorithms within Dynamically Changing Heterogeneous Environments
Nikolay Alekseevich Butakov, Denis A. Nasonov, Alexander V. Boukhanovsky · 2014
State-of-the-art distributed computational environments requires increasingly flexible and efficient workflow scheduling procedures in order to satisfy the increasing requirements of the scientific community. In this paper, we present a novel, nature-inspired scheduling approach based on the leveraging of inherited populations in order to increase the quality of generated planning solutions for the occurrence of system events such as a computational resources crash or a task delay with the rescheduling phase .The proposed approach is based on a hybrid algorithm which was described in our previous work and includes strong points of list-based heuristics and evolutionary meta-heuristics principles. In this paper we also experimentally show that the proposed extension of hybrid algorithms generates more effective solutions than the basic one in dynamically heterogeneous computational changing environments.