Framework DevOps para Projetos de Integração

• Main Author: Azevedo, Jorge Gabriel Fontes Maciel
• Publication Date: 2023
• Format: Master thesis
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10400.22/23431
• topic: DevOps; CI/CD; Agnostic Framework; Integration Services

Deloitte is a company recognized for its services in Consulting and Audit. Integration Services is one of the teams in Consulting at Deloitte and its expertise is developing integration layers for enterprise applications, enabling decoupled communications between them. Some of the common factors in each project are source code management, moving new developments to production, and the developer’s relation with all the remaining teams in the project special the operations team. Historically each project has its own way of doing each of the points mentioned before. These processes are tightly integrated with the factors around the projects when defining it, such as infrastructure, team composition, team experience, etc... As consulting professionals, Deloitte employees usually change between projects which means they need to learn these project-specific processes when they join a new project, and this is a very time-consuming activity until they are completely comfortable. The goal of this project is to create a standard framework, with its own processes, tools, and rules for these activities. The intent is to reduce the strain on the consultants when changing between projects, with the added benefit of having an asset that can evolve with time. It will be refined in each iteration and built with the knowledge of many professionals. All of this is to be able to adapt to a wider scope of requirements Deloitte can benefit by reusing it. The customers will also receive a much more complete product they can rely on even in more extreme circumstances like migrating the infrastructure where the applications are running. It was developed as a Proof of Concept for the project, and when compared with the previous implementations, it improved the development implementation time in new projects. Was identified a performance penalty of around 3 minutes in each CI/CD pipeline execution. It was suggested to switch the Proof of Concept implementation from Groovy to Python in order to lessen the problem. The team is very interested in the project and the future benefits that can be derived from it.