Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/christophermayes/lume-impact-live-demo
Demonstration of LUME-Impact running a live model fetching data from the LCLS EPICS network
https://github.com/christophermayes/lume-impact-live-demo
Last synced: 25 days ago
JSON representation
Demonstration of LUME-Impact running a live model fetching data from the LCLS EPICS network
- Host: GitHub
- URL: https://github.com/christophermayes/lume-impact-live-demo
- Owner: ChristopherMayes
- License: apache-2.0
- Created: 2021-02-20T00:40:42.000Z (almost 4 years ago)
- Default Branch: main
- Last Pushed: 2023-10-27T20:54:45.000Z (about 1 year ago)
- Last Synced: 2024-12-10T12:03:18.941Z (26 days ago)
- Language: Jupyter Notebook
- Size: 11.5 MB
- Stars: 4
- Watchers: 2
- Forks: 7
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# lume-impact-live-demo
Demonstration of LUME-Impact running a live model fetching data from the LCLS EPICS network
# Setup
Make the `lume-live-dev` environment:
`conda env create --file lume-live-dev.yml`
Convert notebooks to `.py` files:
```bash
./build.bash
```
EPICS:
```bash
ssh -fN -L 24666:lcls-prod01.slac.stanford.edu:5068 centos7.slac.stanford.edu
export EPICS_CA_NAME_SERVERS=localhost:24666
```
Test with: `caget KLYS:LI22:11:KPHR`
# Run
For example, LCLS on SDF with default parameters:
```
ipython lume-impact-live-demo.py
```
LCLS on S3DF with default parameters:
```
ipython lume-impact-live-demo.py -- -t "s3df"
```
# Parameters
All Parameters to run lume-impact-live-demo.py file -
```
-d Debug=True/False
-l Live=True/False
-v USE_VCC=True/False
-m Please pass the model name here (sc_inj/facet/lcls)
-h Please pass the host name here (sdf/singularity)
Defaults -
debug=False
use_vcc=True
live=True
model=sc_inj
host=sdf
```
# Running Lume-Impact-Live-Demo on S3DF
Please remember to follow all the above steps (including connecting to PVs). Please wait roughly 5-10 mins for job to complete one simulation.
```
export LCLS_LATTICE=/sdf/group/ad/beamphysics/lcls-lattice
export SCRATCH=/sdf/group/ad/beamphysics/lume-impact-live-demo/SCRATCH3 (Try not using the lscratch or scratch on S3DF. Somehow, it didn't work for me.)
export LUME_IMPACT_CODEBASE_LOCATION=/sdf/group/ad/beamphysics/lume-impact-live-demo
export LUME_OUTPUT_FOLDERS=/sdf/group/ad/beamphysics/lume-impact-live-demo/output
export SCRATCH=/sdf/group/ad/beamphysics/lume-impact-live-demo/SCRATCH
ipython /sdf/group/ad/beamphysics/lume-impact-live-demo/lume-impact-live-demo.py -- -t "s3df"
```
# Running Lume-Impact-Live-Demo on S3DF as a Service
Easiest way to run lume-impact-live-demo on S3DF as a Service is through Cron Jobs.
Please remember to check all the export variables in `cron_job_setup.sh`
Saving below config should automatically trigger the job in background. Use logs to monitor the job. Otherwise, You can see the pid saved in ~/tmp/myprogram.pid which signifies the last pid that executed this program.
```
crontab -e
#Opens a file. Please enter below details and save it.
SHELL=/bin/bash
BASH_ENV=~/.bashrc
* * * * * source /sdf/group/ad/beamphysics/lume-impact-live-demo/cron_job_setup.sh
```
## TOML properties file
Running the simulation requires definition of the following variables within a toml file:
| Variable | Description |
|---------------------------|--------------------------------------|
| host | Host of Impact sim |
| config_file | Impact configuration file |
| distgen_input_file | Input to distgen generation |
| workdir | Working directory of simulation run |
| summary_output_dir | Output directory for summary files |
| plot_output_dir | Output directory of plot files |
| archive_dir | Output directory for archive files |
| shapshot_dir | Output directory for snapshot files |
| distgen_laser_file | File for generating distgen input |
| num_procs | Number of processes to use |
| mpi_run | command for running mpi |
Running on SDF additionally requires:
| Variable | Description |
|---------------------------|--------------------------------------|
| impact_command | Command for Impact execution |
| impact_command_mpi | Command for running MPI |
Default configurations are given in the example environment files packaged with this repository.