Flink SQL: Custom timestamp-based parallel partition scan

Question

Problem

I have a huge table (44G) that doesn\'t have a self-incrementing primary key. It means I cannot ETL this table using optimization params under scan.pa