You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Nov 20, 2019. It is now read-only.
I can image that throttling for the comments (duplicates) is needed, or where you more thinking in the line of throttling the issue / coment creation. Since different user can use different parts of the application and both would generate a lot of issue when f.i. the database connection is failing but the hash would be different.
F.I:
When the count = X it stop inserted for an hour. After an hour the count expires and another X can be inserted.
Throttling. Keeping track of an count per hash, but the timeout is per hash entry.
I think a PassiveExpiringMap and maybe a Synvar could be helpful but it depends on the feature.
@jorkzijlstra I would definitely use the per hash version. We could adjust the hashing mechanism to use (for example) only a subset of the stacktrace to hash.
I am unsure what the best approach would be. Maybe after the first 10 comments directly and after that throttling, one every X time? Maybe also a global cap / throttle. Maybe just solve it for the use case you are experiencing now and adjust when you have more information on what would be useful.
If you make it configurable with a sensible default, users can tweak (or turnoff) the behavior.
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
No description provided.
The text was updated successfully, but these errors were encountered: