AWS Cloud: Infrastructure, DevOps techniques, State of Art.

This thesis consists in an analysis of a project developed in collaboration with Akka Technologies Company. The implemented software is a back-end service built with the support of Amazon Web Services technology, which is the leading cloud provider service. The purpose of this work is to review the best strategies and technologies used during the development of the project. However, the thesis can’t enter either the specifics code details or its future applications due to the customer required secrecy. During the implementation of the project, the first decision taken is to employ the cloud to host the architecture in order to lower the cost, improve the scalability and reliability. Therefore, the second decision is related to the choice of the best cloud provider. Amazon Web Services (AWS) is a slightly better option than the competitors. The development paradigm used is called serverless, that means the developer does not need to build his own server infrastructure but he can exploit the one furnished by the cloud provider. This is a big advantage in terms of deployment rapidity, cost and easiness of making the software scale because it does not require further investments in hardware or time. In fact, to change the service options, only few actions are needed. AWS offers a long lists of services that can be used to realize cloud projects. After the analysis of the most important ones, the basic services for realizing the project are EKS and Lambda. Lambda is the best choice because its pricing is based on usage and its running time must not overcome 15 minutes which is more than enough to run an API REST that save and retrieve data on the database. The others main architectural components used are the API Gateway that exposes the endpoints necessary to reach the lambda functions, the Secret Manager which keep stored environmental variables, Cognito that is the authentication service and RDS which is the database one. The process of development is realized using the best practices proposed from DevOps. Therefore the app is a micro-services one, in fact each lambda works independently from the others. The job is organized using Scrum to try to keep the a good visibility of the tasks and following an idea of learning culture and improvement of the team members. The techniques of continuous delivery, testing, secure deployment, logging systems, control metrics and cloud security are realized at state of art. Also the Infrastructure as Code paradigm is implemented with Terraform service that let you create a cloud infrastructure only using code.