Microservices: Building Blocks to New Workflows and Virtualization

Abstract

As the media production and broadcast industry continue to transition toward internet protocol and commercial off-the-shelf (COTS) equipment, it is clear there is a looming gap between how nontraditional media companies leverage cloud and datacenter environments and how broadcasters think about those same networks. Companies like Facebook, Google, and Amazon see their networks as being dynamic, living environments that are defined on demand and provisioned at will for scale. The traditional broadcast view would have those same networks more locked down and fixed in functionality, but control over scale is certainly a requirement for future deployments. For broadcasters and other media companies to remain competitive, it is imperative that they transition operations to a more agile and flexible workflow design architecture. This is where microservices, a component-based approach to application and service design, come into play. The Facebooks and Googles of the world have already jumped into the world of microservices, building large-scale, high-performance networks that maximize the efficiency of COTS platforms. The success that those companies have experienced in their move to microservices is undeniable, but how does that apply to the video-centric world and the requirements of nonlinear and live broadcast? Can microservices be leveraged for all aspects of television? This tutorial-type paper will explain what microservices are and how they compare to virtual machines, containers and service-oriented architecture (SOAs), as well as how they potentially interact. This paper will look at platforms and potential hardware, as well as software-based architecture decisions. Finally, this paper will consider how software strategies used in other industries can translate to media production and broadcast, ensuring that the industry can successfully respond to continually changing consumer demand.