Controlling data flow comprises one part of troubleshooting in your replication environment.
The other part is tuning servers and components. You use either or both of the methods based on the problems in the environment.