Improvising Container Scheduling using Genetic Algorithm
In modern software development it is getting increasingly common to replace monolithic software with light-weight services called micro services. A micro services is a cohesive, independent process interacting via messages, where multiple micro services perform distinct tasks independent of each other. Visual Machines (VMs) and software containers enable virtualization and can be used easily to deploy micro services, where resources for VMs and containers (CPU, RAM, and bandwidth) can be set depending on user needs, and needs of the deployed application. Containers emerged as both a new technology for virtualization and for easier application delivery with increased usage of micro services, the popularity of containers has increased even further. This paper proposes new algorithm to improve the container scheduling performance in terms of cloud load balancing and reliability by taking tasks form Data Center Controller (DCC). To achieve this, an improved genetic algorithm is used. This minimizes the make span and utilizes the resources effectively.