Mitigating Resource Management and Continuous Integration Obstacles in Heavy Traffic Systems Using Containerization and Orchestration Tools

Abstract

In this paper I covered virtualization technologies, including Virtual Machines and Containerization engines analyzing their advantages and drawbacks. After reviewing possible container application areas, I introduce Kubernetes, an orchestration tool that manages a cluster of containers on autopilot given correct configurations. After that, I give more details on how an orchestration framework functions, communicates with containers, and keeps the containers up. Moreover, I propose CI/CD tools and compare their efficiency through a designed experiment built using a cluster of nodes provided by the accessible version of Google Cloud Platform and deployed containers via Google Kubernetes Engine. At the end of the experiment, I give a comprehensive analysis of the results. To summarize, both containers and virtual machine technologies allow users to describe and develop their software environments before running them on top of multiple resources in a portable, repeatable manner. With containers, it is possible to construct scalable architecture composed of a large number of services (microservices). Also, the integration of new features can be done more efficiently if deployed in a continuous manner using the proposed CI/CD tools. Nevertheless, there are open questions to be researched, such as how the various tools respond when something goes wrong in the pipelines and the best policies for reverting to previous versions to ensure high availability.