Fabric SQL Engine, add some sort of resource control at the Workspace level

today, we can customize Spark compute at the workspace level, this is very handy as some workspace are more "important" than others for example, Finance Workspace is definitely more critical than a test workspace, another reasons maybe even more important, it is much safer to give access to more people if you know that bad actors can do less damages :)

the current workaround is to have multiple capacities, which is problematic for two reasons

• sometimes all you have is one capacity
• unused compute from different capacities can not be shared