Open schampoux opened 4 weeks ago
Our current logging approach is handled primarily by the log_event
function in folding/utils/logging.py
file. The log_event
function takes in an instance of the class it's being called from and an event
dictionary. The function then prepares the event for logging by calling prepare_event_for_logging
function from neurons/validator.py
.
run_cmd_commands
function into a method of a larger running class, that is able to hold the state of the particular command. event
dict that is being passed through something like event.update(...)
.Store
class The most complicated part of this will be pipelining, which includes how to use the class throughout the codebase
The purpose of this issue is to improve our current logging schema in order to get better feedback on failing
gmx
commands.