I often have project where a few features take a magnitude of order longer to run than most. If I have 10 cores sometimes for hours a single core is running while the other 9 are waiting. This repeats for each scenario. If running multiple scenarios (or all) could be run in a single parallel command.
uo run --feature <path/to/FEATUREFILE.json> --scenario <path/to/SCENARIOFILE.csv, , path/to/SCENARIOFILE2.csv>
Or maybe if no scenario passed in then all are run.
uo run --feature <path/to/FEATUREFILE.json>
My work around for this for now will likely be dividing i in --num-parallel=[i] by number of scenarios, and then kicking off multiple runs at once. This may be much quicker than giving all cores to each scenario and running one at a time
I often have project where a few features take a magnitude of order longer to run than most. If I have 10 cores sometimes for hours a single core is running while the other 9 are waiting. This repeats for each scenario. If running multiple scenarios (or all) could be run in a single parallel command.
uo run --feature <path/to/FEATUREFILE.json> --scenario <path/to/SCENARIOFILE.csv, , path/to/SCENARIOFILE2.csv>Or maybe if no scenario passed in then all are run.
uo run --feature <path/to/FEATUREFILE.json>My work around for this for now will likely be dividing
iin--num-parallel=[i]by number of scenarios, and then kicking off multiple runs at once. This may be much quicker than giving all cores to each scenario and running one at a time