Systems of Systems Engineering - A Goal-Driven Architecture Simulation Approach for Cloud Computing Systems

Systems of Systems Engineering, Cloud Benchmarking, Cloud Simulation, Cloud Elasticity

Large-scale systems of systems (SoS) are becoming too complex for systems engineers to comprehend, develop, and operate successfully. SoS failures are “… traceable to excessive complexity, poor architectural choices, ill-defined processes, non-validated systems engineering practices or lack of experience in applying valid practices” [INCOSE Systems Engineering Vision 2020]. The solution contains the following major elements: (1) Design and build a local cloud benchmarking and simulation lab available to the NSF Net-centric and Cloud Software and Systems I/UCRC (NSF NCSS I/UCRC) members; (2) Create a local test-bed to compare genetic algorithm results to benchmarks and simulation results; (3) Add local lab cloud emulation (to benchmark and collect  resource usage metrics) for Google, Amazon, Azure, and OpenStack cloud-provider tests; (4) Add features to RE-Tools to generate goal XML for automatic simulation input; (5) Locate the next Cloud elasticity constraints; (6) Augment the graphics capabilities of the Cloud Simulation/Forecaster GUI.