Recent advancements in cloud technology have significantly enhanced the deployment and development of innovative technological solutions. Central to this progress are virtual machines optimized for high-performance tasks, utilizing non-uniform memory access (NUMA) technology to ensure consistent and reliable performance through low-latency, localized memory access for each central processing unit. This paper presents a comprehensive test suite specifically designed to evaluate NUMA’s reliability across various stress scenarios. By leveraging cloud-based tools and implementing test case reduction strategies, the suite not only improves execution time efficiency, but also provides a detailed assessment of NUMA’s operational resilience. Our findings demonstrate the effectiveness of the test suite in identifying and managing multiple uncorrectable memory errors, thereby validating its utility in optimizing testing processes. Notably, the application of this suite reduced the execution time of test cases from 24 to 14 h, marking a significant advancement in testing efficiency. This study underscores the potential of integrated NUMA technology in cloud environments to meet the demands of high-performance applications, highlighting substantial improvements in both efficiency and resource management.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Software Test Suite to Stress Non-Uniform Memory Access on Cluster Environments Using Cloud Technologies

  • J. A. Torres Ramos

摘要

Recent advancements in cloud technology have significantly enhanced the deployment and development of innovative technological solutions. Central to this progress are virtual machines optimized for high-performance tasks, utilizing non-uniform memory access (NUMA) technology to ensure consistent and reliable performance through low-latency, localized memory access for each central processing unit. This paper presents a comprehensive test suite specifically designed to evaluate NUMA’s reliability across various stress scenarios. By leveraging cloud-based tools and implementing test case reduction strategies, the suite not only improves execution time efficiency, but also provides a detailed assessment of NUMA’s operational resilience. Our findings demonstrate the effectiveness of the test suite in identifying and managing multiple uncorrectable memory errors, thereby validating its utility in optimizing testing processes. Notably, the application of this suite reduced the execution time of test cases from 24 to 14 h, marking a significant advancement in testing efficiency. This study underscores the potential of integrated NUMA technology in cloud environments to meet the demands of high-performance applications, highlighting substantial improvements in both efficiency and resource management.