A framework upgrade can pass review, pass staging, and still fail in production. The code may be correct in isolation, but production adds the parts staging rarely carries: real traffic, background workers, connection pressure, failover paths, and configuration drift.
That is the trap. A green staging run proves the app works against a smaller copy of the world. It does not prove the system will hold when traffic, concurrency, and operational history arrive at the same time.
For high-risk framework or infrastructure changes, the useful test is production-shaped: realistic load, realistic failure paths, clear rollback criteria, and the right owners online. Anything less is not useless, but it should be treated as confidence-building, not proof.