What causes UNSUPPORTED_DATA_TYPE?

Writing a STRUCT or MAP column to a JDBC sink (e.g. MySQL, SQL Server) that does not support complex types. A Databricks JDBC source returns a database-specific type with no Spark equivalent. Using INTERVAL or VOID types in an operation that requires a concrete primitive type. A Delta Lake table was created with a type supported in a newer Databricks Runtime not available in the current cluster. A connector (e.g. Databricks ODBC driver) does not support a newly introduced Spark type

How do I fix UNSUPPORTED_DATA_TYPE?

Cast the unsupported column to STRING or JSON string before writing: to_json(struct_col) AS col_json.. Use explode or inline to flatten ARRAY or STRUCT columns before JDBC writes.. Check the target system's supported type list and map each Spark column type to a compatible target type.. For JDBC connectors, specify a custom type mapping with the customSchema option.. Upgrade the Databricks Runtime if the type was added in a newer version and you need native support.

Medium severitydata quality

Power BI Refresh Error:
UNSUPPORTED_DATA_TYPE

What does this error mean?

An operation or connector encountered a data type that it does not support. This error typically occurs when writing to an external system that lacks an equivalent type for a Spark SQL type, or when using a Spark operation that does not accept a specific complex type.

Common causes

1Writing a STRUCT or MAP column to a JDBC sink (e.g. MySQL, SQL Server) that does not support complex types
2A Databricks JDBC source returns a database-specific type with no Spark equivalent
3Using INTERVAL or VOID types in an operation that requires a concrete primitive type
4A Delta Lake table was created with a type supported in a newer Databricks Runtime not available in the current cluster
5A connector (e.g. Databricks ODBC driver) does not support a newly introduced Spark type

How to fix it

1Cast the unsupported column to STRING or JSON string before writing: to_json(struct_col) AS col_json.
2Use explode or inline to flatten ARRAY or STRUCT columns before JDBC writes.
3Check the target system's supported type list and map each Spark column type to a compatible target type.
4For JDBC connectors, specify a custom type mapping with the customSchema option.
5Upgrade the Databricks Runtime if the type was added in a newer version and you need native support.

Frequently asked questions

Does this error occur in Delta-to-Delta writes?

Rarely. Delta Lake supports all Spark SQL types. UNSUPPORTED_DATA_TYPE most commonly appears when writing from Delta to external systems (JDBC, ODBC, CSV) that have narrower type support.

What is the best way to handle STRUCT columns in a Postgres write?

Serialize the struct column to a JSON string with to_json(struct_col) and map it to a TEXT or JSONB column in PostgreSQL.