Length problems are common in old systems
Legacy databases often use short fixed-length fields. Modern sources do not always respect those limits. A product description from an API is often longer than the database column waiting for it.
Advanced ETL Processor validates field length before the writer step. The workflow rejects the row, routes it to an exception table, or deliberately trims values when that is safe and agreed.
Fields worth checking
- Customer, supplier, and product names.
- Postcodes, state codes, country codes, and identifiers.
- Email addresses and URLs.
- Free-text fields loaded into legacy databases.
Do not trim blindly
Trimming text is useful when it is a business decision. Cutting an internal note is usually acceptable. Cutting an account reference is how future you develops trust issues.
Useful references
See loading data into MS SQL Server, choosing ETL software, and SQL Server documentation on char and varchar data types.