Characterizing output bottlenecks in a supercomputer
Citations Over TimeTop 10% of 2012 papers
Abstract
Supercomputer I/O loads are often dominated by writes. HPC (High Performance Computing) file systems are designed to absorb these bursty outputs at high bandwidth through massive parallelism. However, the delivered write bandwidth often falls well below the peak. This paper characterizes the data absorption behavior of a center-wide shared Lustre parallel file system on the Jaguar supercomputer. We use a statistical methodology to address the challenges of accurately measuring a shared machine under production load and to obtain the distribution of bandwidth across samples of compute nodes, storage targets, and time intervals. We observe and quantify limitations from competing traffic, contention on storage servers and I/O routers, concurrency limitations in the client compute node operating systems, and the impact of variance (stragglers) on coupled output such as striping. We then examine the implications of our results for application performance and the design of I/O middleware systems on shared supercomputers.
Related Papers
- → Evaluation of active storage strategies for the lustre parallel file system(2007)91 cited
- → Black-box problem diagnosis in parallel file systems(2010)47 cited
- → Characterizing I/O Optimization Effect Through Holistic Log Data Analysis of Parallel File Systems and Interconnects(2020)1 cited
- → Research on the Establishment of Distributed Storage Virtual Geographic Environment Based on Lustre File System(2011)
- Design and Implementation of File-Level RAID Based on Object-Based Storage System(2010)