External File ingestion

[JibinSebastian]

Hi All,

I want to ingest external csv from from a different network to fabric, Could you please  help me to understand the best way to ingest that data into fabric.

any leads would be greatly appreciated.

best regards,
Jibin Sebastian

[ati_puri]

Hi @JibinSebastian ,

Few options suggested for your scenaio:

Use On-Premises Data Gateway

• Install Microsoft On-Premises Data Gateway on your local network.
• Configure Fabric Dataflows (Power Query) or Data Factory pipelines to connect through the gateway.
• This allows Fabric to securely access local files without moving them manually.

2. Scheduled Upload via Automation

• Use Power Automate or a script (PowerShell, Python) to:
  • Pick up CSV files from local folders.
  • Push them to OneLake or Azure Blob Storage.
• Once in OneLake, Fabric can ingest them into Lakehouse or Dataflow Gen2.

3. Hybrid Approach

• If you have network restrictions, you can:
  • Set up a shared folder accessible via gateway.
  • Or use SFTP/FTP with a pipeline in Fabric Data Factory.

Let me know if the requirement is different.

If you found this helpful, do like it and accept it as a solution.

[lbendlin]

Please clarify what "different network" means.  For example - if these CSV files are on a sharepoint then you can use the new shortcut feature - no need to ingest anything.

[deborshi_nag]

Hi @JibinSebastian 

You can use a Dataflows Gen2 using the File connector. Whether you will need a gateway to connect to the file will depend on:

• you need a gateway if:

  • The external CSV file is on a private network (on-premises or VPC) that is not accessible over the public internet.
  • You want secure connectivity without opening ports or exposing the network publicly.

• No gateway needed if:

  • The CSV file is hosted on a publicly accessible endpoint (HTTP, HTTPS).
  • The file is in a cloud storage service (Azure Blob, AWS S3, etc.) with public or token-based access.
  • You manually upload the file into OneLake.

Hope this helps - please appreciate leaving a Kudos or accepting as a Solution!