Currently we store many benchmark-configurations in the benchmarking database which the post-results.py script matches via the --storage-configuration option. We should have our data generation scripts add information to the metadata.json file about how the data were generated and what benchmark configuration they should match; that way we can have the script automatically detect, or otherwise verify, the storage configuration for a given run instead of blindly trusting the user's input.
This will also hopefully help us reduce the number of potential duplicate storage configurations if it is clear how they differ or not.
Currently we store many benchmark-configurations in the benchmarking database which the
post-results.pyscript matches via the--storage-configurationoption. We should have our data generation scripts add information to the metadata.json file about how the data were generated and what benchmark configuration they should match; that way we can have the script automatically detect, or otherwise verify, the storage configuration for a given run instead of blindly trusting the user's input.This will also hopefully help us reduce the number of potential duplicate storage configurations if it is clear how they differ or not.