RECAP Work Package 8 (WP8) is designed to address the challenges and complexities of optimized application placement and infrastructure deployment for the next generation of cloud-edge technologies. This deliverable is the final report that disseminates the technical research for this work package, and presents the final methodologies, algorithms and RECAP use-case specific implementations developed as part of WP8.
The research is encapsulated as an integral RECAP architecture component, termed as the 'Placement & Infrastructure Optimiser', and the deliverable defines each individual step taken to provide the optimized solutions, which includes the:
- Development of generically applicable models that can be easily be applied beyond RECAP to other cloud/edge-based use-cases. These focus on contextualizing the available infrastructure, defining the application load with analogous component migration and load distribution analysis, and translating the application load into infrastructure capacity and utilization metrics, [section 2].
- Design and formulation of optimization algorithms that address the trade-off between multiple stakeholder business objectives, including non-linear functions. A non-dominated sorting genetic algorithm (when considering non-linearity) and a revised evolutionary algorithm (along with the use of utility functions to normalize objectives allowing a comparison across potential solutions) are presented, [section 3].
- Use-Case specific implementations and results for the RECAP use-cases defining the application of the developed models and algorithms to provide optimized solutions to the specific research problem for the use-case, [section 4].
The deliverable, thus, concludes the research work for WP8, successfully demonstrating optimized capacity planning of infrastructure resource configurations and placement of virtual components on physical infrastructure resources across distributed edge and cloud data centres.
To download this deliverable, please fill in the form.