Abstract—I/O has become one of the determining factors of HPC application performance. Understanding an application’s I/O activity requires a multi-level view of the I/O function flow that includes high-level I/O libraries. We have developed a tracing framework, called Recorder, that captures I/O function calls at multiple layers of the parallel I/O stack without requiring source code modifications. In this paper, we show how Recorder’s trace output can be used to investigate I/O activity and identify performance inefficiencies in two I/O benchmarks running on a leading edge HPC platform. Future work to organize and present the collected information more intuitively will further increase the value of Recorder’s capabilities. We believe that...
Using parallel file systems efficiently is a tricky problem due to inter-dependencies among multiple...
In high-performance computing (HPC) environments, an appropriate amount of hardware resources must b...
HPC systems are becoming ever more important as a data processing resource for the LHC experiments. ...
The 2014 TOP500 supercomputer list includes over 40 deployed petascale systems, and the high perform...
Abstract. Efficient execution of parallel scientific applications requires high-performance storage ...
Efficient usage of file systems poses a major challenge for highly scalable parallel applications. T...
The ever-growing CPU-I/O performance gap I/O has become one of the determining factors 2 Figure cite...
Two key changes are driving an immediate need for deeper understanding of I/O workloads in high-perf...
This artifact contains the prototype implementation of a novel approach for recording I/O operations...
Competing workloads on a shared storage system cause I/O resource contention and application perform...
Beside the computational scalability of an HPC application, its I/O behaviour can significantly infl...
The computing power of high-performance computing (HPC) systems is increasing with a rapid growth in...
Competing workloads on a shared storage system cause I/O resource contention and application perform...
I/O efficiency is essential to productivity in scientific computing, especially as many scientific d...
High performance computing (HPC) is changing the way science is performed in the 21st Century; exper...
Using parallel file systems efficiently is a tricky problem due to inter-dependencies among multiple...
In high-performance computing (HPC) environments, an appropriate amount of hardware resources must b...
HPC systems are becoming ever more important as a data processing resource for the LHC experiments. ...
The 2014 TOP500 supercomputer list includes over 40 deployed petascale systems, and the high perform...
Abstract. Efficient execution of parallel scientific applications requires high-performance storage ...
Efficient usage of file systems poses a major challenge for highly scalable parallel applications. T...
The ever-growing CPU-I/O performance gap I/O has become one of the determining factors 2 Figure cite...
Two key changes are driving an immediate need for deeper understanding of I/O workloads in high-perf...
This artifact contains the prototype implementation of a novel approach for recording I/O operations...
Competing workloads on a shared storage system cause I/O resource contention and application perform...
Beside the computational scalability of an HPC application, its I/O behaviour can significantly infl...
The computing power of high-performance computing (HPC) systems is increasing with a rapid growth in...
Competing workloads on a shared storage system cause I/O resource contention and application perform...
I/O efficiency is essential to productivity in scientific computing, especially as many scientific d...
High performance computing (HPC) is changing the way science is performed in the 21st Century; exper...
Using parallel file systems efficiently is a tricky problem due to inter-dependencies among multiple...
In high-performance computing (HPC) environments, an appropriate amount of hardware resources must b...
HPC systems are becoming ever more important as a data processing resource for the LHC experiments. ...