Forum Discussion
10gb limit reached
Hi everyone
We have a situation in our environment where we're butting up against the 10gb limit in our premium capacity workspace. We are operating a F32 fabric reservation with 3 workspaces (dev/test/prod), we have a semantic model (cube) that we've had to cut down in terms of history as it exceeded 10gb when we processed over the weekend - are there any options available to us to increase the capacity of the workspace so we can bring back the desired history - this is a non negotiable end user requirement.
Apologies for the 'laymans' explanation, but grateful for any advice.
Many thanks
JF
Hi JAFpwrBI_002 ,
As lbendlin mentioned, you need at least P1 or A4 SKUs to support a single semantic model larger than 10GB. P1 or A4 allows a single semantic model to be up to 25GB.
What is Power BI Premium? - Power BI | Microsoft Learn
Capacity and SKUs in Power BI embedded analytics - Power BI | Microsoft Learn
Since your semantic model is larger than 10GB, this means that the full refresh operation uses up to twice the memory required by the model alone. Incremental refreshes reduce memory and improve refresh performance by partitioned refreshes of tables.Incremental refresh for semantic models in Power BI - Power BI | Microsoft Learn
Thanks for lbendlin and ibarrau responses and contributions to the community.
Best regards,
Mengmeng Li
5 Replies
Hi. Even though there are settings that can help, I would start with a deep analysis for downsizing and improving details at the semantic model. You can check 5 tips in here: https://blog.ladataweb.com.ar/post/614108758904389632/datamodeling-5-tips-de-reducción-de-tamaño-de
I'm saying that because after working with large amount of data, when the data model and the granularity was established by a strong process, it usually won't size that much.
After saying that you can check if large storage format setting could take it. That enables you to prevent the 10gb limit of workspace, but I'm not sure that will make the report work properly. It will depend on the design of the semantic model and the frequency of the end users getting daily details.
More info: https://learn.microsoft.com/en-us/power-bi/enterprise/service-premium-large-modelsConsder incremental refresh for the solution
If the limit is reached you can think about Power Bi Aggregation. It allows you to import a smaller granularity than 10gb and only if the user query needs more details goes through direct query to get the data, making the dataset smaller.
I hope that helps,
Remember the 10GB limit is per partition. Not per semantic model or workspace.
Use incremental refresh. (You may need to upgrade to a F64 to enable that)
Hi, at the moment we're doing full reloads but plan to move to incremental in the future - how does that help with the space issue?
With incremental refresh your semantic model size can be (much) larger than 10GB. Your Capacity type will define how much larger. A F64 will give you up to 25GB, for example. Ish.
Hi JAFpwrBI_002 ,
As lbendlin mentioned, you need at least P1 or A4 SKUs to support a single semantic model larger than 10GB. P1 or A4 allows a single semantic model to be up to 25GB.
What is Power BI Premium? - Power BI | Microsoft Learn
Capacity and SKUs in Power BI embedded analytics - Power BI | Microsoft Learn
Since your semantic model is larger than 10GB, this means that the full refresh operation uses up to twice the memory required by the model alone. Incremental refreshes reduce memory and improve refresh performance by partitioned refreshes of tables.Incremental refresh for semantic models in Power BI - Power BI | Microsoft Learn
Thanks for lbendlin and ibarrau responses and contributions to the community.
Best regards,
Mengmeng Li
