Loading...

Model-Driven Approach for Creating Language-Agnostic Models and Automated Generation of IaC

Zamiri, Negar | 2025

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 58240 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Ramsin, Raman
  7. Abstract:
  8. Infrastructure management (including the setup of networks, virtual machines, servers, and software stacks) is an essential activity in cloud environments and data centers. Infrastructure as Code, or IaC for short, refers to the process of managing the infrastructure using a high-level programming language. It offers numerous benefits, such as cost reduction, increased speed, and decreased risk and human intervention. However, IaC is currently afflicted with serious problems, including tedious, manual and error-prone tasks, along with requiring the users to create IaC solutions through scripting, which calls for comprehensive knowledge and training across a wide range of diverse and heterogeneous cloud platforms. A model-driven engineering approach can facilitate the creation of complex code and tools, as well as help build language-independent models that are automatically translated into IaC. In this thesis, a model-driven engineering approach was adopted to facilitate the automatic generation of IaC code from abstract, language-independent models. Initially, existing methods, tools, and patterns in the IaC domain were thoroughly reviewed and analyzed to identify key challenges and requirements. Based on this analysis, appropriate modeling levels and transformation rules were designed with a focus on the specific characteristics of IaC. Accordingly, a tool was developed to perform transformations from high-level abstract models to executable code, along with a chatbot assistant that receives high-level model information in natural language. The proposed solution was applied in practical projects within two real-world organizations, and its performance was evaluated based on defined criteria. The results demonstrated that the proposed approach contributes to reducing complexity, improving development speed, decreasing the risk of human error, and easing the use of IaC for end users
  9. Keywords:
  10. Model Driven Approach ; Automatic Code Generation ; Infrastructure as Code (IaC) ; Models Transformation

 Digital Object List

 Bookmark

No TOC