A Novel Container Placement Mechanism Based on Whale Optimization Algorithm for CaaS Clouds

Abstract

Advancements in container technology can improve the efficiency of cloud systems by reducing the initiation time of virtual machines (VMs) and improving portability. Therefore, many cloud service providers offer cloud services based on the container as a service (CaaS) model. Container placement (CP) is a mechanism that allocates containers to a pool of VMs by mapping new containers to VMs and simultaneously considering VM placements on physical machines. The CP mechanism can serve several purposes, such as reducing power consumption and optimizing resource availability. This study presents directed container placement (DCP), a novel policy for placing containers in CaaS cloud systems. DCP extends the whale optimization algorithm, an optimization technique aimed at reducing the power consumption in cloud systems with a minimum effect on the overall performance. The proposed mechanism is evaluated against established methods, namely, improved genetic algorithm and discrete whale optimization using two criteria: energy savings and search time. The experiments demonstrate that DCP consumes approximately 78% less power and reduces the search time by approximately 50% in homogeneous clouds. In addition, DCP saves power by approximately 85% and reduces the search time by approximately 30% in heterogeneous clouds.