Reaching capacity limit - Unable to load model due to reaching limits

[amir_mm]

Could someone please help me understand this SKU and CU capacities?

We are on A2 capacity (with 5 GB memory) and today the reports were not loading and the error was: "Unable to load model due to reaching limits" , so had to move to A3.

Is this 5GB memory per month? what could cause to reach the limit? Does increasing the number of refreshing could cause this issue? 

As you see below, we were close to 100% CU in some days but we touched that limit today. Does it mean that we can get close to 5GB anytime but as soon as we hit the 100% limit we are forced to upgrade SKU to another level?

[Anonymous · Accepted solution]

Hi @amir_mm ,

1. SKU (Stock Keeping Unit):

   • In Microsoft Fabric, a SKU represents a specific resource tier with varying levels of memory and computing power.
   • Each SKU is associated with a certain amount of Capacity Units (CUs), which measure the compute power available for that SKU.
   • The type of SKU you need depends on the specific solution you want to deploy.

2. Capacity Units (CU):

   • CU is a unit of measurement for compute power within a SKU.
   • It quantifies the resources available for a given SKU.
   • Different SKUs provide different CU values, reflecting varying levels of computational capacity.

3. Is the 5GB memory per month?

   • No, the 5 GB memory is not per month. It represents the available memory within your A2 capacity.
   • This memory is accessible as long as you stay within the capacity limits.

4. What could cause reaching the limit?

   • Several factors can contribute to reaching capacity limits:
     • Data volume: Loading large datasets or complex reports can consume memory and processing power.
     • Concurrent users: If many users access reports simultaneously, it can strain the capacity.
     • Complex calculations: Resource-intensive calculations may impact capacity.
     • Data refresh frequency: Frequent data refreshes can consume resources.
     • Queries and visuals: Complex queries or visuals may require more memory.

5. Does increasing the number of refreshes cause issues?

   • Frequent data refreshes can indeed impact capacity.
   • Each refresh consumes resources (memory and processing).
   • Consider optimizing refresh schedules or using incremental refresh where possible.

As you said, when your capacity reaches 100% CU utilization, it means that you have fully utilized the available computing power. You can upgrade to a higher SKU at this point: migrate from A2 to A3 to provide more resources or keep a close eye on CU utilization and plan your upgrade accordingly.

Below is the official link will help you:

Capacity and SKUs in Power BI embedded analytics - Power BI | Microsoft Learn

Best Regards,

Xianda Tang

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[GilbertQ]

Hi @amir_mm 

You're actually limited more by CPU than memory and that is why you're going over the CU limited I would suggest looking at your semantic models to ensure that they are using best practise to consume as little CPU as possible.