DIAL: 通过学习的客户端本地度量值实现去中心化的I/O自动调优以平行文件系统
DIAL: Decentralized I/O AutoTuning via Learned Client-side Local Metrics for Parallel File System
AI总结:
DIAL通过学习客户端本地度量值实现去中心化的I/O自动调优,提升并行文件系统的性能。
摘要:
在当今高性能计算系统中,使并行文件系统(PFS)实现高效、高性能的数据访问至关重要。PFS客户端侧I/O对最终交付到个别应用程序和整个系统的I/O性能有重大影响。对关键客户端侧I/O行为进行自动调优已得到广泛研究,并显示出有希望的结果。然而,现有工作严重依赖大量的全局运行时度量来监控和准确建模应用程序的I/O模式。这种大量的开销显著限制了在实际系统中实现细粒度、动态调优的能力。在本研究中,我们提出了DIAL(通过学习的客户端本地度量值实现去中心化的I/O自动调优),其采取了截然不同的方法。与其尝试提取应用程序的全局I/O模式,DIAL采取去中心化的方法,将每个I/O客户端视为独立单元,并仅使用其本地可观察的度量来调整配置。借助机器学习模型,DIAL使多个可调单元能够做出独立但集体的决策,及时响应全球存储系统中发生的情况,并在全局范围内为应用程序实现更好的I/O性能。
英文摘要:
Enabling efficient, high-performance data access in parallel file systems (PFS) is critical for today's high-performance computing systems. PFS client-side I/O heavily impacts the final I/O performance delivered to individual applications and the entire system. Autotuning the key client-side I/O behaviors has been extensively studied and shows promising results. However, existing work has heavily relied on extensive number of global runtime metrics to monitor and accurate modeling of applications' I/O patterns. Such heavy overheads significantly limit the ability to enable fine-grained, dynamic tuning in practical systems. In this study, we propose DIAL (Decentralized I/O AutoTuning via Learned Client-side Local Metrics) which takes a drastically different approach. Instead of trying to extract the global I/O patterns of applications, DIAL takes a decentralized approach, treating each I/O client as an independent unit and tuning configurations using only its locally observable metrics. With the help of machine learning models, DIAL enables multiple tunable units to make independent but collective decisions, reacting to what is happening in the global storage systems in a timely manner and achieving better I/O performance globally for the application.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。