A Reference Architecture for On-Premises Software-Defined Infrastructure Using Model-Based Systems Engineering

IT leaders have long refined the use of automation of common support processes in the public cloud through development and operations (DevOps) methods and tools, particularly in organizations where custom application development is a core capability. Cloud service providers inherently expose the application programming interfaces (APIs) to support automation, and development organizations have mature processes and tools to support process automation. This is not the case for organizations that heavily depend on commercial off-the-shelf (COTS) applications running in on-premises data centers, particularly healthcare provider organizations. It remains an open question whether these organizations should pursue automation based on value: are the benefits of automation greater than the costs to implement? In order to begin answering the second part of that question, this paper proposes a reference architecture (RA) for software-defined infrastructure (SDI) focused on leveraging infrastructure and management tools in the on-premises data center, using the methods and tools of Model-Based Systems Engineering (MBSE) through the Model-Based System Architecture Process (MBSAP).