Energy-Aware Scientific Real-Time DAG Scheduling Using Slack Reclamation in Heterogeneous DVFS Cloud Datacenters

Provision of a high level of Quality of Services (QOS) to end users in a cost-effective way with minimum power consumption and minimum Service Level Agreement (SLA) violation is one of the main challenges facing cloud service providers. Accordingly, this paper presents an energy-efficient scheduling strategy in SLA-Aware and cost-effective manner for real-time scheduling of parallel tasks modeled as Directed Acyclic Graphs (DAG) in cloud environment. The proposed approach utilizes Dynamic Voltage and Frequency Scaling (DVFS) capability which has been incorporated into recent commodity multi-core processors to fill the gaps within schedule. The aim is to develop an energy-aware strategy which, besides yielding accurate results, maintains the high quality of the delivered services in a standard and financially rational manner. The proposed scheduling algorithm is compared with other basic strategies in terms of different service quality requirements. The performed experiments demonstrate that the proposed approach outperforms existing methods and can be vowed as a promising scheduling strategy. Thirdly, a novel algorithm to find the best combination of frequencies to result the optimal energy is presented. The proposed algorithm has been evaluated based on results obtained from experiments with three different sets of task graphs: 3000 randomly generated task graphs, and 600 task graphs for two popular applications (Gauss–Jordan and Lower and Upper (LU) decomposition). The results show the superiority of the proposed algorithm in comparison with other techniques.