A high-assurance measurement repository system
Farokh Bastani, Simeon C. Ntafos, I‐Ling Yen, David E. Harris, Rich Morrow, R. Paul · 2002
High-quality measurement data are very useful for assessing the efficacy of high-assurance system engineering techniques and tools. Given the rapidly evolving suite of modern tools and techniques, it is helpful to have a large repository of up-to-date measurement data that can be used to quantitatively assess the impact of state-of-the-art techniques on the quality of the resulting systems. For many types of defects, including Y2K failures, infinite loops, memory overflow, access violations, arithmetic overflow, divide-by-zero, off-by-one errors, timing errors, deadlocks, etc., it may be possible to combine data from a large number of projects and use these to make statistical inferences. This paper presents a highly secure and reliable measurement repository system for measurement data acquisition, storage and analysis. The system is being used by the QuEST Forum, which is a new industry forum consisting of over 100 leading telecommunications companies. The paper describes the decisions that were made in the design of the measurement repository system, as well as implementation strategies that were used in achieving a high-level of confidence in the security and reliability of the system.