Forum Discussion
Helper VFabric pipeline error
Hi,
Removing Int32/Int64 columns in your copy activity does not help because dataverse formats all attributes of the virtual entity before sending them out. This is a virtual table plugin/provider issue. The backend data source is returning a decimal/numeric value that is too large or too small for Int64 and dataverse chokes. This is not something Fabric can fix, fabric simply receives whatever dataverse sends.
Hi Hussein_charif,
Thank you Vinodh247 and tayloramy for the responses to the query.
In addition to the previous suggestions, you can try to pinpoint where the overflow is happening on the Dataverse side. Since the error occurs before Fabric accesses the data, the next step is to check the values returned by the virtual table provider. Try limiting the records such as by date or ID range to find which rows cause the issue. Once you’ve identified the problematic rows, review the relevant numeric or calculated fields in the external source, as one of them might be producing a value outside the allowed range.
If you find the field causing the problem, you can modify the mapping in the virtual table provider, perhaps by reducing precision, lowering the scale, or converting the value to text if it’s only for display. If you don’t manage the provider, the Dataverse or external system team may need to create a safer view with properly constrained numeric types and update the virtual table to use that view. Adjusting the mapping at the source is the best way to ensure Fabric pipelines can read the table without errors.
Thank you.
Community SupportHi Hussein_charif,
Thank you Vinodh247 and tayloramy for the responses to the query.
In addition to the previous suggestions, you can try to pinpoint where the overflow is happening on the Dataverse side. Since the error occurs before Fabric accesses the data, the next step is to check the values returned by the virtual table provider. Try limiting the records such as by date or ID range to find which rows cause the issue. Once you’ve identified the problematic rows, review the relevant numeric or calculated fields in the external source, as one of them might be producing a value outside the allowed range.
If you find the field causing the problem, you can modify the mapping in the virtual table provider, perhaps by reducing precision, lowering the scale, or converting the value to text if it’s only for display. If you don’t manage the provider, the Dataverse or external system team may need to create a safer view with properly constrained numeric types and update the virtual table to use that view. Adjusting the mapping at the source is the best way to ensure Fabric pipelines can read the table without errors.
Thank you.