Ultimately, we should only have errors that are actual bugs in Sentry. Errors we can do nothing about, or error that originate from the user (ex: wrong passhrase) should never be raised in Sentry. That way we reduce noise, and it makes every sentry error and error to address.
However the urgency of the fix depends on the context.
Task
We need to identify the errors, and the threshold that would raise an alert.
Identify all possible errors, in frontend and backend.
Categorize them and identify the category, and the threshold that would lead to an alert
Assign the right developers based on the location of the error: backend, frontend, both ?
Ultimately, we should only have errors that are actual bugs in Sentry. Errors we can do nothing about, or error that originate from the user (ex: wrong passhrase) should never be raised in Sentry. That way we reduce noise, and it makes every sentry error and error to address.
However the urgency of the fix depends on the context.
Task
We need to identify the errors, and the threshold that would raise an alert.