Time-discretization for speeding-up scheduling of deadline-constrained workflows in clouds

In this paper we deal with the problem of scheduling workflow applications in multiple infrastructure-as-a-service (IaaS) providers, where the scheduler must determine on which computational resource each component of a workflow should be scheduled in an attempt to minimize the monetary cost of the workflow execution. In our previous work, we describe the λ-granularity approach that increases the degree of discrete time intervals in an integer linear programming (ILP)to speed up the scheduling of workflow applications with deadline constraints. To assist the selection of a specific discrete time interval (λ value) given a workflow and its deadline, we propose in this paper an aggregate objective function which is derived from historical evaluations of scheduling costs and scheduler runtimes. Results of simulations evince that the proposed AOF-based procedure was able to calculate a specific λ value that allowed the reduction of the ILP-based scheduler running time, although yet achieving cost-effective solutions.