|build| Overview ======== Recorder is a multi-level I/O tracing framework that can capture I/O function calls at multiple levels of the I/O stack, including HDF5, PnetCDF, NetCDF, MPI-IO, and POSIX I/O. Recorder requires no modification or recompilation of the application and users can control what levels are traced. Publications ------------ Recorder has undergone significant changes since the last "Recorder 2.0" paper. We have incorporated a new pattern-based compression algorithm, along with many new features. We have submitted a journal paper describing all these changes in detail. The draft version is available on `arXiv `__. `Wang, Chen, Jinghan Sun, Marc Snir, Kathryn Mohror, and Elsa Gonsiorowski. “Recorder 2.0: Efficient Parallel I/O Tracing and Analysis.” In IEEE International Workshop on High-Performance Storage (HPS), 2020. `__ I/O Patterns and consistency requirements ----------------------------------------- `Traces from 17 HPC applications `__ The traces were collected using an old version of Recorder. The current version uses a different trace format. To read those traces please use Recorder 2.2.1 from the `release `__ page. The following paper presents a comprehensive study on the I/O patterns of these applications. Additionally, it investigates the consistency requirements of these applications, confirming experimentally that POSIX consistency is rarely required by HPC applications. `Wang, Chen, Kathryn Mohror, and Marc Snir. “File System Semantics Requirements of HPC Applications.” Proceedings of the 30th International Symposium on High-Performance Parallel and Distributed Computing (HPDC). 2021. `__ .. toctree:: :hidden quickstart build usage features postprocessing changes .. |build| image:: https://github.com/uiuc-hpc/Recorder/actions/workflows/cmake.yml/badge.svg :target: https://github.com/uiuc-hpc/Recorder/actions/workflows/cmake.yml