During an on-premises to AWS migration using AWS MGN (Application Migration Service), the replication is stuck at 99% for the last few GB. The source server is a Linux database server with a large InnoDB redo log. What is the most likely cause?