Detecting I/O Access Patterns of HPC Workloads at Runtime

• Jean Luca Bez UFRGS
• Francieli Zanon Boito University Grenoble Alpes
• Ramon Nou Barcelona Supercomputing Center
• Alberto Miranda Barcelona Supercomputing Center
• Toni Cortes Barcelona Supercomputing Center / Universitat Politécnica de Catalunya
• Philippe O. A. Navaux UFRGS

Resumo

In this paper, we seek to guide optimization and tuning strategies by identifying the application's I/O access pattern. We evaluate three machine learning techniques to automatically detect the I/O access pattern of HPC applications at runtime: decision trees, random forests, and neural networks. We focus on the detection using metrics from file-level accesses as seen by the clients, I/O nodes, and parallel file system servers. We evaluated these detection strategies in a case study in which the accurate detection of the current access pattern is fundamental to adjust a parameter of an I/O scheduling algorithm. We demonstrate that such approaches correctly classify the access pattern, regarding file layout and spatiality of accesses - into the most common ones used by the community and by I/O benchmarking tools to test new I/O optimization - with up to 99% precision. Furthermore, when applied to our study case, it guides a tuning mechanism to achieve 99% of the performance of an Oracle solution.

Palavras-chave: Measurement, Runtime, Optimization, Decision trees, Layout, Servers, Benchmark testing, high-performance computing, parallel I/O, access pattern detection, I/O forwarding, classification