Fix worker to load multiple challenges in dev env #2277
Conversation
| @@ -38,8 +39,8 @@ | |||
| "DJANGO_SETTINGS_MODULE", "settings.dev" | |||
| ) | |||
| DJANGO_SERVER = os.environ.get("DJANGO_SERVER", "localhost") | |||
| LIMIT_CONCURRENT_SUBMISSION_PROCESSING = os.environ.get( | |||
| "LIMIT_CONCURRENT_SUBMISSION_PROCESSING" | |||
| LIMIT_CONCURRENT_SUBMISSION_PROCESSING = eval( | |||
There was a problem hiding this comment.
Instead of eval(), it would definitely be a better idea to do int() since this is always going to be integer. By doing eval(), you are opening doors for bugs later in the codebase.
| if isinstance(challenge_module, ModuleType): | ||
| return | ||
| else: | ||
| sys.exit(1) |
There was a problem hiding this comment.
Why exiting the process? So basically what will happen here is that if the module is not loaded then the process will exit without showing any error. This doesn't sound right to me.
| @@ -249,6 +250,7 @@ def extract_challenge_data(challenge, phases): | |||
| CHALLENGE_IMPORT_STRING.format(challenge_id=challenge.id) | |||
| ) | |||
| EVALUATION_SCRIPTS[challenge.id] = challenge_module | |||
| return challenge_module | |||
There was a problem hiding this comment.
Sorry but I don't think I understand exactly how returning the challenge module here will help solve the problem of loading multiple modules. Can you please elaborate more on how this is solving the problem?
There was a problem hiding this comment.
That was a temporary hack but I found a permanent fix for this while loading the modules dunamically: https://docs.python.org/3/library/importlib.html#importlib.invalidate_caches
|
Closing this PR. |
This PR fixes -
CHALLENGE_PKfrom docker.env file to load multiple challenges in a single worker for dev env.