Currently the post_results.py script takes many parameters that are input manually for the user to post details about a given benchmark run, this includes the storage-configuration, the commits the images were built with, the harware SKU, etc...
This process is unfortunately quite prone to failure, since the user can easily input an option that does not match what the benchmark was actually doing. For instance, there is no way for the script to verify if the storage-configuration that the user claims they ran with is what the benchmark actually ran on.
We should be able to automate the inputs for many of these options (or at least automate verification)
Currently the
post_results.pyscript takes many parameters that are input manually for the user to post details about a given benchmark run, this includes the storage-configuration, the commits the images were built with, the harware SKU, etc...This process is unfortunately quite prone to failure, since the user can easily input an option that does not match what the benchmark was actually doing. For instance, there is no way for the script to verify if the storage-configuration that the user claims they ran with is what the benchmark actually ran on.
We should be able to automate the inputs for many of these options (or at least automate verification)