An Energy-aware Dynamic Scheduling Algorithm for Optimizing Workflows Under Budget-constraints

The provision of cloud computing offers an untapping scalability and elasticity which is best suited for the execution of user tasks and complicated scientific workflows. Regardless, the big problem of workflow scheduling under a user-specified budget still prevails as a result of the task inter dependencies and the resource diversity. This research proposes a hybrid Energy-Aware Enhanced Salp Swarm Algorithm (EA-ESSA), designed to dynamically schedule tasks while adhering to user specified budget constraints. This technique supports dynamic scheduling using task duplication in idle time spots and integrates APIs for real-time spot pricing. This proposed technique also minimizes makespan and energy consumption by improving resource utilization. The algorithm's performance was exhaustively experimented with using both simulated workloads and actual HPC2N datasets. The simulation results show significant advancements in the makespan, resource utilization and energy consumption compared to existing algorithms like ACO, GA, PSO, and MOTSWAO. This research benefits cloud environments comprising complex, unpredictable workflows by cutting environmental effects and shrinking processing expenses.