SinergyCloud: A simulator for evaluation of energy consumption in data centers and hybrid clouds

Abstract

One of the significant limitations in the evaluation of hybrid clouds is the difficulty in validating new solutions by their deployment in real systems since replication of tests on large environments is usually highly expensive. To overcome these limitations, simulators have become popular for conducting preliminary tests. However, there is still a lack of simulation tools with easy-to-use code that allows handling diverse cloud scenarios and the simulation of energy consumption of all devices (hosts, switches, routers, and storage). To fill this gap, the simulator SinergyCloud has been developed to evaluate data centers in hybrid clouds. SinergyCloud allows evaluating diverse cloud scenarios, including energy consumption, workflow makespan, the completion time of tasks, and migrations of virtual machines, with a fine granularity of abstraction. SinergyCloud can handle the simulation of hybrid clouds with multiple data centers composed of hundreds of thousands of devices. Moreover, it is a Java-based, event-driven, and packet-level simulator, having a less steep learning curve than do other simulators. To show the feasibility of SinergyCloud, we performed accuracy and scalability analyses. Metrics values obtained by another simulator had less than 1% difference compared to SinergyCloud, demonstrating its accuracy. In terms of scalability, a scenario with 10,000 hosts was simulated in about 7 h using a typical personal computer. A comprehensive analysis of algorithms for the scheduling of virtual machines in a hybrid cloud is also presented, showing how to perform various evaluations.