Skip to content

NeuwirthLab/NeuwirthLab-ERT4IO-PDSW

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Extended ER4IO (AWS-S3) – PDSW

To better understand the I/O behavior of emerging workloads, provide a more comprehensive characterization of HPC systems, and enable a consistent scoring model in the future, we have adopted the classic Roofline model for I/O workload analysis. Our model offers a clear view of how close observed I/O performance is to peak performance and can also help identify performance bottlenecks.

ERT4IO (https://github.com/NeuwirthLab/ERT4IO) is a Python script that plots the I/O Roofline graph for applications and benchmarks. It was originally based on parsed text files from Darshan outputs. In this work, we extend ERT4IO to also parse data from different output formats.

Usage

python roofline.py

Make sure you have the following Python packages installed:

  • Python ≥ 3.8
  • numpy
  • pandas
  • matplotlib
  • matplotlib-label-line

Cite as

If you use ERT4IO in your research or publications, please cite the following paper:

@inproceedings{Zhu_2023,
  author={Zhu, Zhaobin and Bartelheimer, Niklas and Neuwirth, Sarah},
  title={An Empirical Roofline Model for Extreme-Scale I/O Workload Analysis}, 
  year={2023},
  doi={10.1109/IPDPSW59300.2023.00106},
  booktitle={2023 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)}, 
  pages={622-627},
  keywords={Measurement;Analytical models;Adaptation models;Visualization;Distributed processing;Conferences;Computational modeling;I/O;HPC;Performance Modeling Workflow;I/O Analysis;I/O Performance;I/O System Evaluation}
}

@inproceedings{Tang_2025,
  author = {Tang, Meng and Zhu, Zhaobin and Guo, Luanzheng and Bandy, James G. and Carlson, Tim and Neuwirth, Sarah and Kougkas, Anthony and Sun, Xian-He and Tallent, Nathan R.},
  title = {Quantifying AWS S3 I/O Performance Boundaries Using the Roofline Model},
  year = {2025},
  doi = {10.1145/3731599.3767513},
  booktitle = {Proceedings of the SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis},
  pages = {1415–1423},
  keywords = {AWS S3, distributed workflows, performance analysis, storage bottlenecks, I/O monitoring, I/O roofline model},
  series = {SC Workshops '25}
}

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages