Flush and Fill Verses Incremental Loading
A great place to begin when designing an ETL process is to decide if you will use an incremental load versus the simpler flush and fill technique. With incremental loading, you insert, update, or delete data in destination tables that have changed in the source table on a row by row basis.
Flush and fill is the opposite of incremental loading. Flush and fill completely refreshes the data, by first removing all of the existing data in a destination table and then refilling that table with the source’s current data.
Once you know the advantages and disadvantages to both, you may choose to use one technique on some tables, and the other on the rest
Note: Microsoft is the owner of the content and video in this blog post.