Forum Discussion
Data read from On-prem SQL server to Fabric Satawarehouse is slower through Dataflow GEN2
Memorable Memberjlakshminaraya Also, I got a question. Do you use one or two chained dataflows to do your ingestion from the on-prem DB to the synapse warehouse?
I ask because I am unable to ingest anything from an on-prem DB via the Power BI data gateway to a Fabric lakehouse or warehouse if I only use one dataflow. I have to use the workaround Microsoft recommends. This workaround consist of using 2 dataflows. The first one connects to the on-prem DB and ingest the data. And the second dataflow uses the first dataflow as a source and loads the data to a Fabric destination.
I have one dataflow only. Which is working firn for me
- Element115
That's amazing! Why do I need to use 2 DFs??? I don't get it. Did you also set your firewall rules to this or do you have different endpoints?
- Protocol: TCP
- Endpoints: *.datawarehouse.pbidedicated.windows.net, *.datawarehouse.fabric.microsoft.com, *.dfs.fabric.microsoft.com
- Port: 1433
That's straight from the documentation... but if the docs are not in sync with some new changes, then that might explmain perhaps I am missing an endpoint?
- Element115
jlakshminaraya Could you please let me know what your endpoints are in your firewall rules?
This?
- Endpoints: *.datawarehouse.pbidedicated.windows.net, *.datawarehouse.fabric.microsoft.com, *.dfs.fabric.microsoft.com
Or something else? like an additional endpoint perhaps that's not listed in the docs?
