What causes Fabric Copy Activity Error 20748?

The 'allErrorFile' skipErrorFile option was added to a copy activity that is not configured for binary copy mode. The copy activity operates on individual files or non-folder paths rather than a folder-level binary copy. A pipeline was migrated or cloned from a binary copy setup and the skipErrorFile settings were not updated

How do I fix Fabric Copy Activity Error 20748?

Step 1: Open the copy activity in your Fabric Data Factory pipeline editor and review the format and dataset settings to confirm whether binary copy with folder is configured.. Step 2: If binary copy with folder is not the intended mode, navigate to the copy activity's advanced settings or JSON payload and locate the 'skipErrorFile' block.. Step 3: Remove the 'allErrorFile' property from the skipErrorFile configuration.. Step 4: If binary copy with folder is the intended mode, verify that both source and sink datasets point to folder paths and that the copy activity format is set to binary.. Step 5: Save and republish the pipeline, then re-run to confirm the error is resolved.

Medium severityconfigurationMicrosoft Fabric →

Power BI Refresh Error:
Fabric Copy Activity Error 20748

What does this error mean?

The 'allErrorFile' option in skipErrorFile is only supported when using binary copy with a folder-level configuration. This setting must be removed when those conditions are not met.

Common causes

1The 'allErrorFile' skipErrorFile option was added to a copy activity that is not configured for binary copy mode
2The copy activity operates on individual files or non-folder paths rather than a folder-level binary copy
3A pipeline was migrated or cloned from a binary copy setup and the skipErrorFile settings were not updated

How to fix it

1Step 1: Open the copy activity in your Fabric Data Factory pipeline editor and review the format and dataset settings to confirm whether binary copy with folder is configured.
2Step 2: If binary copy with folder is not the intended mode, navigate to the copy activity's advanced settings or JSON payload and locate the 'skipErrorFile' block.
3Step 3: Remove the 'allErrorFile' property from the skipErrorFile configuration.
4Step 4: If binary copy with folder is the intended mode, verify that both source and sink datasets point to folder paths and that the copy activity format is set to binary.
5Step 5: Save and republish the pipeline, then re-run to confirm the error is resolved.

Frequently asked questions

What exactly is 'binary copy with folder' and how do I enable it?

Binary copy with folder means your copy activity uses a binary format (no schema transformation) and both source and sink datasets are configured to point to a folder path rather than a specific file. In Fabric Data Factory, set the dataset format to Binary and ensure the file path in the dataset resolves to a folder.

Can I use 'allErrorFile' alongside 'invalidFileName' in the same skipErrorFile block?

Both options have distinct support requirements. 'allErrorFile' requires binary copy with folder, while 'invalidFileName' requires direct binary copy with folder. You must ensure your copy activity configuration satisfies the requirements of every skipErrorFile option you include.