ICSA-Lite

Abstract: The need to operate at scale, providing fast and stable performance levels at expected cost, is a common requirement for modern software systems. These systems must leverage sophisticated, highly configurable execution and storage platforms that are designed from the ground up to exploit large scale distributed computational resources. Once architects select preferred execution platforms, applications must be designed and configured to provide required performance and reliability. This requires navigating a complex design space of architectural trade offs and dependent parameter settings. A mature collection of performance prediction and modeling methods and tools are available for architects to utilize to address this problem. In reality however, they are rarely used in practice. This talk will examine the reasons for this lack of uptake of research results by exploring the architectural features and parameter spaces of two widely deployed scalable software platforms. I’ll also describe practical engineering approaches that can provide rapid insights on performance options and ensure systems can provide stable response times and throughput.
Ian Gorton is the director of computer science master’s programs and a professor of the practice at the Khoury College of Computer Sciences at Northeastern University’s Seattle campus. He is passionate about analyzing and designing complex, high-performance scalable distributed systems and embodying design and architecture principles in methods and tools that can be exploited by architects in other projects. Before joining Northeastern in 2015, he was a senior member of the technical staff at the Carnegie Mellon University Software Engineering Institute. He had several projects on the principles of designing massively scalable software architectures for big data applications and building knowledge bases both manually and using machine learning to support engineering tasks. Before this, Gorton was a laboratory fellow in computational sciences and math at Pacific Northwest National Laboratory (PNNL). He managed the Data-Intensive Scientific Computing research group and was the chief architect for PNNL’s Data Intensive Computing Initiative. He was also PI for multiple projects in environmental modeling, carbon capture and sequestration, and bioinformatics. This experience has led to a particular interest in the design of large-scale, highly customizable cyber-infrastructures for scientific research. Gorton is a senior member of the IEEE Computer Society and a Fellow of the Australian Computer Society. Until July 2006, he led the software architecture R&D at National ICT Australia in Sydney, Australia, and previously worked at CSIRO, IBM, Microsoft, and in academia in Australia. A complete list of his publications and citations can be found on Google Scholar and in dblp. He is the author of Essential Software Architecture (2011) and The Foundations of Scalable Systems (2022).

Abstract: In the past decade, we have been fortunate enough to work with several of the most prominent companies in the world to develop and validate our methodology for design analysis, for the purpose of improving software quality and productivity. In this talk, I will share our observations on how practitioners create and maintain architecture models, which presents a significant gap between what we have been teaching in class. After decades of research and education, the idea of software modeling is still not widely accepted. The key challenges include the difficulty of identifying which part of design should be modeled, and most importantly, how to evolve the model as the software evolves. Effective modeling also requires sufficient knowledge on quality attributes, and rallies on the organization’s management process and development culture. The lack of effective software modeling also contributes to the emerging and growth of technical debt. In this talk, I will share our recent experience of re-creating design models for a software product that needs to be refactored and our suggestions on how to improve software modeling in education and in practice.
Dr. Yuanfang Cai is a Professor at Drexel University, and currently working at Google as a visiting researcher during her sabbatical. Her primary research interests are software design, software architecture, software evolution, technical debt detection, and software economics. She has created a number of influential concepts, methods, and tools, including Design Rule Space, Decoupling Level, Modularity Violation, Hotspot patterns, etc. Her team has created and validated the automated detection and quantification of design debt, supported by their DV8 tool suite, on hundreds of open source projects and dozens of industrial projects from various companies. She has authored about 100 publications and four patents. Dr. Cai is currently serving on program committees and organizing committees for multiple top conferences, and serves as an Associate Editor and editorial board member for top journals in the area of software engineering. The tools and technologies from Dr Cai’s research have been licensed and adopted by multiple multinational corporations.