OneLake shortcuts: permissions and security boundaries for zero-copy data access
Shortcuts expose data across workspaces without copying it, but identity flow and permissions on both the shortcut and target paths must be designed.
A shortcut is a pointer, not a copy
A OneLake shortcut makes data in a target location visible from another path without creating a new physical copy. This can reduce duplication and synchronization work. But avoiding a copy does not make access automatically secure. Evaluate permissions on both the shortcut path and the target path.
Permissions on both paths work together
Creating a shortcut requires write permission on the destination item and read access to the target data. When a user reads through the shortcut, permissions on the shortcut path and target path combine; the more restrictive access applies. Granting access only in the consumer workspace does not remove source-side access requirements. Test the permission matrix for each user role.
Passthrough and delegated identity
With passthrough, the caller's identity is passed to the target and source-side permissions remain in control. With delegated identity, a configured connection identity or credential is used. Shortcuts to external cloud stores may require delegated authentication. The two patterns differ in auditability and user experience; do not choose one only because setup is easier.
A Direct Lake identity edge case
Some query paths using Direct Lake on SQL or delegated identity can access target data with the item owner's identity rather than the end user's identity. OneLake security roles may still apply, but direct target permissions for the end user might not behave as expected. This matters for sensitive finance, HR, and customer data. Run end-to-end access tests with real user roles.
A production readiness checklist
Record the data owner, source path, consumer path, identity mode, permission scope, and data classification for every shortcut. Test Admin, Member, Contributor, and Viewer roles separately. For external stores, test credential rotation and access revocation. Document that deleting a shortcut removes the reference, not the source data.
Validate platform choices against real workloads
Choose Warehouse, Lakehouse, Eventhouse or another processing layer from workload requirements: query patterns, event latency, file sizes, data quality and access controls. Build a representative pilot and compare at least two suitable architectures using the same criteria. Document update, deletion, versioning and retention semantics before treating a demonstration as production-ready.
Manage capacity and ownership together
Analytics cost includes compute, refresh frequency, failed-job retries, replicated models and access control, not only storage volume. Track cost alongside data freshness and user-facing latency. Assign an owner, service target, schema-change policy and incident process to each dataset. Scale capacity only after evidence identifies the actual bottleneck.
Comments
No comments yet. Be the first to contribute.
Try the comment interactions · demo preview
The two examples below are only a UI demo, not real visitor comments. Likes and replies are local to this page and are not saved; use the real form below to post.
I'd like to see which steps benefit most in real deployment.
An example test scenario would make the technical analysis clearer.
✎ Write a comment