Research Area:  Cloud Computing
For a cloud datacenter, the biggest issue is how to handle millions of requests coming dynamically from the end users. To distribute requests efficiently and effectively, there is a need to balance the load evenly among the Cloud federations. In this paper, we propose a load balancing strategy and apply it on two levels control in a datacenter: physicals machines and clusters. The algorithm proposed is based on hierarchical strategy. Simulations on CloudSim are used to check its performance and its influence on standard deviation of load, communication cost and number of task migrations. Comparing with an algorithm, that does not include different levels in load balancing decision; results indicate that our method not only gives good Cloud balancing but also ensures optimizing Standard Deviation of load, resources utilization and communication cost by reducing number of migrations in all the datacenter.
Author(s) Name:  Amal ZaouchFaouzia BenabbouNaoufal Er-Raji
Conferrence name:  mation and Communication Conference FICC 2021: Advances in Information and Communication
Publisher name:  Springer
Volume Information:  Volume 2021
Paper Link:   https://link.springer.com/chapter/10.1007/978-3-030-73103-8_20