Hypervisor-based hardware virtualization- also known as the first phase of virtualization uses Virtual Machines (VM) to provide better hardware resource utilization and application isolation. A VM provides some level of portability, but still requires a full operating system (OS) with all the binaries and libraries required to run the service it hosts. Therefore, moving an application from a development to a production environment for instance is no different than moving them between two Physical Machines (PM). Container-based virtualization-sometimes known as the next phase of virtualization addresses some of these limitations by providing virtualization at the OS level. Docker is an open source engine launched in 2013 by a company called Docker, Inc. Docker is used to manage the lifecycle of containers. Using containers, it is no longer necessary to dedicate an entire VM to an application in order to provide isolation, thus saving OS license costs. In this project we plan to formulate a generic model that can be used to fine tune a container-based setup for maximum performance benefit.
