This project focused on developing automated techniques to enhance the efficiency and reliability of distributed systems, essential for modern applications like cloud computing and big data processing. As these systems grow more complex, ensuring their smooth and efficient operation is critical.

Intellectual Merit:

The intellectual merit of this project lies in developing novel automated techniques to refactor distributed systems. Refactoring refers to improving the internal structure of code without altering its external behavior. In distributed systems, refactoring is challenging due to asynchronous communication and non-deterministic behaviors arising from the interaction of multiple components across machines.

Our work addressed these challenges by creating new methods for integrating, testing, and optimizing distributed systems. A key innovation was enabling client applications to leverage edge resources, enhancing performance through the replication of cloud services on edge devices. We also developed a tool for testing replicated data libraries, helping to identify and resolve issues caused by complex message exchanges. Additionally, a study on web services provided valuable insights into factors influencing their reliability and popularity. Lastly, we developed a lightweight caching architecture for Wi-Fi access points, reducing latency in time-sensitive applications like AR/VR and autonomous driving.

These contributions advance the understanding and development of distributed systems, offering practical solutions to improve their performance, consistency, and robustness.

Broader Impacts:

This project extends its impact beyond technical contributions. Distributed systems are integral to modern computing, supporting a wide range of services from cloud storage to online applications. By improving their reliability and efficiency through automated refactoring, the project helps ensure these systems remain robust and scalable, particularly as they handle increasingly large and complex datasets.

The findings also have educational value. The PI has incorporated key concepts into courses, giving students exposure to cutting-edge software development techniques in distributed systems through hands-on learning.

Additionally, the PI has initiated discussions with an industry partner to explore future research opportunities and technology transfer. This collaboration connects academic research with real-world applications, paving the way for future innovations and partnerships.

Summary of Outcomes:

This project developed innovative techniques to improve the efficiency and reliability of distributed systems. Key achievements include creating a system that allows client applications to leverage edge resources by replicating cloud services on edge devices. We also introduced an approach for testing replicated data libraries and conducted a study on web services to identify factors affecting their reliability. Additionally, a lightweight caching architecture for Wi-Fi access points was developed, reducing latency for time-sensitive applications.

These contributions enhance the understanding of distributed systems, providing practical solutions for improving performance, consistency, and robustness. The project also had educational impacts, integrating key concepts into coursework, and initiated discussions with an industry partner for future research and technology transfer.