Common errors when loading data into a cloud data warehouse via Etlworks, with fixes. The issues below have been observed across Snowflake, Amazon Redshift, Google BigQuery, Azure Synapse Analytics, Microsoft Fabric Warehouse, Databricks, Vertica, and Greenplum — they're caused by how the warehouse parses staged CSV files, not by any one warehouse's quirks.
Column type 'VALUE' is not recognized
Example: Timestamp WEB is not recognized.
The error means the staged file's columns don't line up with the destination table — either the column order or the inferred data types are wrong.
The order of columns in the file doesn't match the table
Fix. Enable Reorder columns to load to match the order of columns in the target table. Also update column types to match the target under MAPPING → Parameters → Handling schema changes.
The column data type in the file doesn't match the table
If Etlworks auto-created the destination table, it may have picked the wrong type. For example, an alphanumeric column inferred as NUMBER (or NUMERIC, INT64, …) will fail on the load with the "Column type is not recognized" error.
Fix. Two options on the CSV format used by the flow:
- Enable Save Metadata so source types propagate to the destination table.
- Enable All fields are strings — all destination columns get a TEXT / VARCHAR / STRING type, side-stepping type-inference issues.
End of the record reached while expected to parse column "COLUMN NAME"
Special characters in the data are breaking the warehouse's CSV parser.
Fix. Update the CSV format used by the flow:
- Set Enclosure Character to ".
- Enable Always enclose.
- Enable Escape double-quotes.