What is a common mistake when working with distributed tracing and correlation IDs?

A common mistake with distributed tracing and correlation IDs is generating a new unrelated identifier in every service breaks the end-to-end chain and makes troubleshooting distributed failures much harder. The safer approach is to design and test the behavior explicitly rather than assuming the platform or dependency will handle it automatically.