Add more Fabric work without adding more pipelines
More tables, Lakehouses and Warehouses bring more dependencies and more ways for a run to fail halfway through. Weaver keeps using the same project files and the same commands as the work grows.
More objects, the same way of working
Weaver keeps working out the order
Add Python, SQL or another Fabric item. Weaver reads the new dependencies from the code, so you do not need another set of pipelines to connect the new work to the old.
Change one part without rebuilding everything
Weaver rebuilds the selected objects and the downstream work affected by them. It leaves independent parts of the project alone.
Rerun the work that needs it
When one part fails, you should not have to rerun everything.
Continue from the last clean load
Weaver records how far each object loaded, so incremental work can resume from a known position.
Keep independent work running
If one object fails, its dependants wait. Other branches can continue when you choose fault-tolerant loading.
See what still needs attention
Health reports what failed, what was blocked and what became stale after an upstream load.
Branch the data as well as the code
A Git branch gives you another version of the project files. Weaver Mirror gives you a development branch of the Fabric data too.
The development project can start from current production data without copying every table. Lakehouse data can remain behind OneLake shortcuts and Warehouse data behind views.
Build the parts you need to change in development. Leave the rest mirrored until they need to diverge.