UCERF3-ETAS Measurements
Page is under active construction, some sections may be incomplete. Aug 13 2024 - bhatthal@usc.edu
This page summarizes the performance study of UCERF3-ETAS ran locally in Docker and on CARC Discovery, SDSC Expanse, and TACC Frontera. This study allows for evaluation of resource requirements in single-node and multiple-node simulations of the Ridgecrest M7.1 ETAS forecast (ci38457511).
Contents
Installation and Configuration
Running ETAS simulations on OpenSHA is simplified through a collection of launcher binaries and scripts called ucerf3-etas-launcher. The process of installation and configuration varies across systems, however the foundations remain the same. Running simulations always occurs in three phases.
- Building configuration files for a specified event, where we configure MPI nodes and number of simulations
- Launching the simulation with the configuration files
- Consolidating and plotting simulations data
Docker
TODO(bhatthal)
- Explain Docker installation, DockerHub and Dockerfile set up
- Docker resource provisioning
- Access Jupyter notebook terminal via web client
- Config generation with u3etas_comcat_event_config_builder. Explain hpc-site and slurm file vs config.json
- u3etas_launcher.sh and plot generator
Discovery
TODO(bhatthal)
- Explain config generation with slurm_interactive.sh. Reiterate why this is necessary and Head vs compute nodes. Add existing resources to reference for more information.
- Discuss how bashrc and slurm files had to be modified to accomodate this system for both single-node and multiple-node simulations.
- slurm_submit.sh for simulation and plotting
Expanse
TODO(bhatthal)
Frontera
TODO(bhatthal)
Performance Results
TODO(bhatthal)
Docker
TODO(bhatthal)
Discovery
TODO(bhatthal)
Expanse
TODO(bhatthal)
Frontera
TODO(bhatthal)
Conclusion
TODO(bhatthal)
