Autoenabled StrawberryIntegration cost me an hour of downtime - maybe reconsider this behavior?

[ipeterov]

1. Rewrote my app's API from graphene-django to Strawberry (without touching my sentry_sdk.init( call which didn't have any integrations)
2. Tested thoroughly locally - API tests, Playwright tests, manually checked everything just to be sure
3. Deployed and watched everything break down with There is no current event loop in thread 'ThreadPoolExecutor-28_0'. exceptions on every request to my API

Turns out, Sentry somehow detected that I was using Strawberry, but didn't realize I was using GraphQLView and not AsyncGraphQLView. I'm assuming it started running async code, and that broke because it was doing it in a sync context.

I got really confused - even launching the production Docker image locally was completely fine. I just started adding the same build args, and it broke when I added SENTRY_DSN. Then, I found this issue in the Strawberry repo. Turns out, I'm not the first one.

[sentrivana]

Hey @ipeterov, thanks for reporting, I made an issue out of this #4980 -- we'll take a look. 🙏🏻