FWIW, I think there's another (arguably worse) version of this bug where onPreResponse recovers from the error with a 200 response.

I'm not sure it's valuable for the tests to enumerate all these possibilities though. We're just relying on Hapi to report an error on the error channel when applicable. The details of that are not Sentry's responsibility. But also I'm not a maintainer so don't mind me!

@camsteffen Thanks for you engagement in this topic! 👍🏻 I am not sure if I understood you correctly. I added a test case for a "2xx override" inside onPreResponse. In our opinion, this should not be reported to Sentry as this could be done on purpose.

I guess what I'm suggesting is - I wonder if you could have a test like does not report errors from Hapi events on "internal" channel rather than trying to have a test for every possible scenario of an error appearing on the internal channel.