The examples on how reservations are applied are a bit strange.

[simonlid]

In the section Discount examples the examples of how they are applied is on the consumed consumption units and not on the running capacities. This is a bit confusing since to my knowledge you are paying for the capacity if it is in the running state regardless if the CU's are consumed or not. I think that it would be better to reword the example 1 to something like the following:

You purchase a Fabric capacity reservation of 64 CUs. You deploy a F32 capacity used for running Power BI and you also deploy another F32 capacity used for Synapse Data Warehouse. In this case, both capacities get reservation discounts. No usage is charged using pay-as-you-go rates.

Example 2 becomes even more confusing I do understand the concept of smoothing but I think that this example is very confusing. Because if you would to beyond your deployed capacity even with smoothing then interactive delays and rejection logic would kick in. Why are we even including this as an example.

I would also like to see an example of how it would work if you make a capacity reservation with 64 CU's and you deploy one F32 capacity, two F16 capacities and one F8 capacity at the same time, so consuming 72 CU's at the same time. From my understanding you would pay for 64 CU's at the reservation price and for 8 CU's at the pay-as-you-go rate.

---

Document Details

⚠ Do not edit this section. It is required for learn.microsoft.com ➟ GitHub issue linking.

• ID: 6b5c1c87-6850-d459-9312-5fb2b01e5e97
• Version Independent ID: 48bc06bf-b7e1-9088-c886-e8324f1613da
• Content: Save costs with Microsoft Fabric Capacity reservations - Microsoft Cost Management
• Content Source: articles/cost-management-billing/reservations/fabric-capacity.md
• Service: cost-management-billing
• Sub-service: reservations
• GitHub Login: @bandersmsft
• Microsoft Alias: banders

[Naveenommi-MSFT]

@simonlid Thanks for your feedback! We will investigate and update as appropriate.

[bandersmsft]

@simonlid Sorry the delay here. I spoke with the Fabric enginnering team I've updated the examples based on their feedback.

#please-close