In this session, we strive to deliver more insights into the working of Fabric Capacities and the abstraction of Capacity Units. What pragmatic ways are there to predict if a solution is relatively low or high in the usage of Capacity Units? How do we monitor the usage and how can we resolve problems quickly in case of emergency?

Topics:

• Funny (hopefully) real-life story on how one user blew up the entire production Fabric Capacity, which resulted in errors on datasets and reports.

• What lessons can learned from this

• What is the concept of the Microsoft Fabric Capacity and what does bursting mean

• What are capacity units and how are these measured per different resource

• What are ‘expensive/cheap’ resources and patterns

• How does an end-user request translate to the usage of capacity units

• How can you monitor the usage per capacity and resource

• What measures can you take to separate different workloads and different environments (lifecycle)