I’ve got a 3.6 billion row fact table in Azure SQL which I am trying to create a copy of. Initially I setup a simple copy activity in Azure Data Factory to copy from table 1 to table 2. The throughput of this was very poor, under 2MB per second, and this is without any indexing applied.

I later copied to data into Blob Storage, then back into Azure SQL to help with the I/O but this doesn’t improve anything.

It’s a very big SQL Server, and I am seeing that it doesn’t exceed 7% resource utilization for anything which makes me think something is throttling the throughput.

Can someone suggest ways to identify what this might be?