DLH.io does not handle deletions for structural source system metadata. For example, if a customer has a source like a PostgreSQL database and synchronizes a set of database tables to a target, and then at the source a column is dropped, DLH.io will not remove the column from the target system in subsequent data loads.
This is by design to protect data integrity. The downstream impact of automatically removing a target column that previously contained data used for production reporting could be severe. If a column is dropped in the source system, the target column will remain but will no longer be updated with data by DLH.io.