Development and analysis of FinOps processes for a Multi-Cloud environment

Cloud Computing has led companies to rely on remote services instead of build-up their own datacenters. Nowadays many companies are switching to multi-cloud environments to be able to apply a methodological and technical approach to use various cloud based services in a coherent and integrated modality. However, the bigger is the company, the harder it is to control cloud costs. An innovative approach that can solve this problem is FinOps: a methodology which introduces financial practices and experts into Development and Operational teams. This thesis work will focus on researching the limitations of FinOps in its current state, developing new proactive and reactive processes that can extend the FinOps practices into the Infrastructure as Code (IaC) approach. After the development and implementation, there will be also an experimental analysis phase regarding these processes, to validate them. All the developed processes will be built through: Jenkins, an important automation tool, via the creation of scripted automation or CI/CD pipelines; Pulumi, an innovative IaC tool, via the development of Policy as Code scripts; custom scripts written in Python that interact with the Cloud Service Providers through their native APIs. The processes will extend over multiple cloud providers, so that users will interact with only one system, performing multiple operations. The Cloud Service Providers that will be taken in account are Amazon Web Services (AWS) and Microsoft Azure. The aim of the primary process developed is the proactive automation and verification of resource tagging, to avoid human errors. The other processes will help in the management of a multi-cloud infrastructure via the creation and the control of per project budgets, and help in the creation of periodic reports after a total check of all the deployed resources with the possibility of performing corrective actions, as destroying some resources which violate the budget or the policy standard. The experimental validation will verify the correct application and results of the processes developed via statistical comparison between companies that would use these solutions with companies at the actual state. Moreover, a use case study will be presented to explain the correct application of these processes. The thesis structure is composed by an introduction, the background information, the analysis of the FinOps problem in a multi-cloud environment, the development and explanation of the proactive and reactive processes, the experimental validation, a use case study and the conclusions. Moreover, the thesis work provides a solid foundation for future studies into the application of FinOps practices in the multi-cloud world, with the possibility of extending it to additional cloud providers, integrating the developed pipelines with existing CI/CD pipelines for IaC or adding other features to this solution.