Currently, when calling a submicroflow, it always runs within the same transaction unless we use a task queue. I propose adding an option to call a submicroflow in a separate transaction (without relying on task queues).
The benefit of this approach is that if an error occurs in the submicroflow, and we have error handling configured (e.g., Custom with Rollback), then only the changes made within that submicroflow should be rolled back. This would allow us to revert specific changes without affecting the entire main transaction.
There are scenarios where we want to undo certain operations but not all. For example, deleting data after using Custom without Rollback is not always feasible. Having the ability to isolate submicroflow transactions would give us much more flexibility in managing data consistency.
Suggested Implementation: When calling a submicroflow, there could be a checkbox or setting to specify whether it should run in a separate transaction. This would allow developers to easily control whether the submicroflow is part of the main transaction or handled independently.
This feature would be extremely useful in cases where fine-grained transaction management is required, helping us better balance rollback behavior between main flows and subflows.