MetricsIngest is a package in the LSST Science Pipelines.
The program creates scripts to be used to creat metrics files and inject results to squash. It also can create step yaml files for test-med-1 production. To make it flexible for use by different users and for different stacks user need to create and modify minput.yaml file. User also need to get squash login and password. This can be received through dm-hsc-reprocessing slack channel asking Angelo Fausti - @afausti.
- dataset: Data used for test
- 'DC2_test-med-1'
- run_id: Current ticket
- 'DM-36611'
- version_tag: Current stack version
- 'w_2022_42'
- sand_box: Directory where json files will be created
- '/sdf/group/rubin/sandbox/kuropatk/metrics_json/'
- date_tag: Date when production was complete
- '2022-10-20T07:51'
- squash_login: Squash login
- '<squash login>'
- squash_pas: Squash password
- '<squash password>'
- repo: Repository where production results are located
- '/sdf/group/rubin/repo/dc2/'
- run: Path to the production results
- '2.2i/runs/test-med-1/w_2022_42/DM-36611'
setup -r . Metrics
metrics
| --help, -h | : Help on commands |
make-dispatch-sl : Create dispatch.sl script
make-dispatch-sh : Create dispatch.sh script
make-js-file : Create make_json.sl script
make-steps : Create step yaml files for step1 - step9
metrics make-steps ./minput.yaml
metrics make-js-file ./minput.yaml
sbatch make_json.sl
sacct - to see running jobs scancel JobId - to cancel some job if need
metrics make-dispatch-sl ./minput.yaml
metrics make-dispatch-sh ./minput.yaml
chmod +x dispatch.sh
sbatch dispatch.sl
The last command will run for a long time.