What causes DF-Executor-InvalidInputColumns?

The source table or file added, renamed, or removed columns since the data flow was created, and the source projection was not refreshed. Schema drift is disabled on the source transformation, and new columns arriving in the source are rejected as invalid inputs. The source dataset points to a different table or file than when the schema was imported, and the schemas differ. Column names from a JSON or XML source have changed namespace or case since the schema was imported

How do I fix DF-Executor-InvalidInputColumns?

Open the failing source transformation in ADF Studio and click 'Import projection' to refresh the column list from the current source schema.. Check the ADF activity run output for the specific column names flagged as invalid inputs.. If columns were renamed or removed in the source, update all transformations that reference the old names.. Enable schema drift on the source transformation to allow the data flow to handle source schema changes without breaking.. Enable debug mode and preview the source to confirm the actual columns being ingested match what the downstream transformations expect.

Medium severitydata flow

Power BI Refresh Error:
DF-Executor-InvalidInputColumns

What does this error mean?

One or more input columns are invalid — column names in the source or incoming data stream do not match what the transformation expects. The source schema changed (columns renamed, removed, or reordered) and the data flow projection is now out of sync with the actual source.

Common causes

1The source table or file added, renamed, or removed columns since the data flow was created, and the source projection was not refreshed
2Schema drift is disabled on the source transformation, and new columns arriving in the source are rejected as invalid inputs
3The source dataset points to a different table or file than when the schema was imported, and the schemas differ
4Column names from a JSON or XML source have changed namespace or case since the schema was imported

How to fix it

1Open the failing source transformation in ADF Studio and click 'Import projection' to refresh the column list from the current source schema.
2Check the ADF activity run output for the specific column names flagged as invalid inputs.
3If columns were renamed or removed in the source, update all transformations that reference the old names.
4Enable schema drift on the source transformation to allow the data flow to handle source schema changes without breaking.
5Enable debug mode and preview the source to confirm the actual columns being ingested match what the downstream transformations expect.

Frequently asked questions

What is the difference between InvalidInputColumns and ColumnNotFound?

InvalidInputColumns means the entire source projection is out of sync with the actual schema. ColumnNotFound is more specific — a named column in an expression doesn't exist. InvalidInputColumns can be triggered by multiple invalid names or types simultaneously.

How do I refresh the source projection after a schema change?

In the source transformation, go to Projection → 'Import projection'. After refreshing, update downstream transformations (Derived Column, Select, sink mapping) that reference columns by old names.

Does enabling schema drift fix InvalidInputColumns?

Schema drift allows new or unexpected columns to pass through, preventing InvalidInputColumns from added columns. If the issue is a renamed column referenced by the old name in downstream expressions, you still need to update those expressions.

Will downstream Power BI datasets be affected?

Yes — the pipeline fails and the target table receives no new data. Dependent datasets serve stale figures.

Official documentation: https://learn.microsoft.com/en-us/azure/data-factory/data-flow-troubleshoot-guide