We are trying to implement a new solution where we use a Synapse pipeline to copy data from an Azure Databricks Delta Lake to a Synapse dedicated pool. I'm using a copy activity where the source is a dynamically constructed query to be executed against Databricks. The sink is an existing table in the dedicated pool. Copy method is set to Polybase with staging enabled in the settings.
When I run the pipeline I get the following error:
Operation failed: "The specified filesystem does not exist." 404, HEAD.
However, when I copy/paste the exact same query in the expression field using Preview data it runs and gives me the expected output.
We are using the unity catalog in Databricks and I'm worried this may not yet have been properly implemented in Synapse, but I don't understand why the preview would work but the debug does not. Any help would be much appreciated.