Forum Discussion
External data sharing in fabric
Hi Sudhavanikolla ,
We’ve been working with external data sharing across tenants as well, and a lot of the behaviour you’re asking about comes down to how Azure handles network traffic behind the scenes. Fabric sits on top of Azure networking, so the same rules apply.
When you create a shortcut from a Lakehouse in Tenant 1 to a Lakehouse in Tenant 2, Fabric basically has to read the underlying files from the source region. If the tenants are in different regions (for example US → India), Azure treats that as cross-region data egress, and yes, that does come with a cost. The exact price depends on the Azure region pair, but the general rule is:
Moving data out of a region is charged, and the receiving tenant doesn’t pay — only the region where the data originates.If both tenants are in the same region, the cost is usually much lower or sometimes not charged at all, because it stays inside the same Azure regional boundary. When they’re in different countries or completely different regions, the cost is higher. Azure publishes up-to-date pricing, but it varies too often to give a single number here — the safest thing is to check the Azure bandwidth pricing page for your specific region pair.
As for limitations, the shortcut mechanism itself doesn’t really limit you — you can share large amounts of data — but performance depends on the distance between regions. Cross-continent shortcuts will naturally be slower because the files have to be read across long network paths.
On the security side, Microsoft handles the compliance part through standard Azure controls:
– Entra ID identities
– token-based access
– encryption in transit and at rest
– and the fact that shortcuts never copy data by themselves, they only reference itOnly the identities you explicitly allow can read the data, even if the shortcut exists in another tenant.
For the last part of your question: when new data lands in the source Lakehouse, the shortcut doesn’t copy anything on its own — it simply “points” to the folder in Tenant 1. If something in Tenant 2 reads that folder again, Azure will charge only for the amount of data actually read, not the full dataset unless you read all of it again. So if your downstream logic only queries incremental data, you only pay for that incremental egress. If your job reads everything from scratch, then you pay for everything again. It all depends on how much data your processing pipeline touches.
Hope this helps. If so, please give a Kudos 👍 or mark as Accepted Solution ✔️
Gopi Krishna
Hi Sudhavanikolla ,
We’ve been working with external data sharing across tenants as well, and a lot of the behaviour you’re asking about comes down to how Azure handles network traffic behind the scenes. Fabric sits on top of Azure networking, so the same rules apply.
When you create a shortcut from a Lakehouse in Tenant 1 to a Lakehouse in Tenant 2, Fabric basically has to read the underlying files from the source region. If the tenants are in different regions (for example US → India), Azure treats that as cross-region data egress, and yes, that does come with a cost. The exact price depends on the Azure region pair, but the general rule is:
Moving data out of a region is charged, and the receiving tenant doesn’t pay — only the region where the data originates.
If both tenants are in the same region, the cost is usually much lower or sometimes not charged at all, because it stays inside the same Azure regional boundary. When they’re in different countries or completely different regions, the cost is higher. Azure publishes up-to-date pricing, but it varies too often to give a single number here — the safest thing is to check the Azure bandwidth pricing page for your specific region pair.
As for limitations, the shortcut mechanism itself doesn’t really limit you — you can share large amounts of data — but performance depends on the distance between regions. Cross-continent shortcuts will naturally be slower because the files have to be read across long network paths.
On the security side, Microsoft handles the compliance part through standard Azure controls:
– Entra ID identities
– token-based access
– encryption in transit and at rest
– and the fact that shortcuts never copy data by themselves, they only reference it
Only the identities you explicitly allow can read the data, even if the shortcut exists in another tenant.
For the last part of your question: when new data lands in the source Lakehouse, the shortcut doesn’t copy anything on its own — it simply “points” to the folder in Tenant 1. If something in Tenant 2 reads that folder again, Azure will charge only for the amount of data actually read, not the full dataset unless you read all of it again. So if your downstream logic only queries incremental data, you only pay for that incremental egress. If your job reads everything from scratch, then you pay for everything again. It all depends on how much data your processing pipeline touches.
Hope this helps. If so, please give a Kudos 👍 or mark as Accepted Solution ✔️
Gopi Krishna