Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Be one of the first to start using Fabric Databases. View on-demand sessions with database experts and the Microsoft product team to learn just how easy it is to get started. Watch now

Reply
albertoserinf
Regular Visitor

Power BI Premium per user and 10Gb limit for a single table on gateway

I was wondering if with the new license model "Premium per user" we will avoid to incur in this error message we sometimes get during refresh via on-prem gateway:

"The amount of uncompressed data on the gateway client has exceeded the limit of 10 GB for a single table. Please consider reducing the use of highly repetitive strings values through normalized keys, removing unused columns, or upgrading to Power BI Premium".

I see that the model size limit increases up to 100GB with this new license, but does this restriction still apply?

Thanks

Regards

Alberto

1 ACCEPTED SOLUTION

@AbhiSSRS your reply, unfortunatley, is incorrect. The limit you are referencing (100gb per dataset) is not relevant to what the OP is actually asking: they are hitting the limit that is imposed upon tables in the datasource -- with Pro and PPU, there is a limit when refreshing through the gateway of 10gb per table. This is a totally different limit than the 100gb dataset maximum that you are referencing. 

 

Pro and PPU accounts both used shared capacities, and the 10gb-per-table on gatewayrefresh limit applies to all shared capacities. The only way to avoid this limit is with a Premium capacity, as it is not a shared capacity.

 

@albertoserinf I'd be interested to hear what solution you eventually came up with to avoid encountering the same error message. I imagine you were forced to modify your table-size in the database itself.

View solution in original post

5 REPLIES 5
v-robertq-msft
Community Support
Community Support

Hi, @albertoserinf 

According to your description, I think the restriction still apply when you upgraded to the PPU, because the model size limit of PPU is 100 GB per user, but not 100GB per dataset.

v-robertq-msft_0-1620810763000.png

 

Therefore, I think reducing the use of highly repetitive strings values through normalized keys and removing unused columns can be the more useful way for you to avoid this problem.

https://docs.microsoft.com/en-us/power-bi/admin/service-premium-per-user-faq

 

What’s more, you can also consider turning on the large dataset storage format in the premium capacity to make the dataset size limited by the Premium capacity size or the maximum size set by the administrator if it’s possible.

v-robertq-msft_1-1620810763002.png

 

https://docs.microsoft.com/en-us/power-bi/admin/service-premium-large-models

 

Thank you very much!

 

Best Regards,

Community Support Team _Robert Qin

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

AbhiSSRS
Solution Sage
Solution Sage

<removing the incorrect response>

@AbhiSSRS your reply, unfortunatley, is incorrect. The limit you are referencing (100gb per dataset) is not relevant to what the OP is actually asking: they are hitting the limit that is imposed upon tables in the datasource -- with Pro and PPU, there is a limit when refreshing through the gateway of 10gb per table. This is a totally different limit than the 100gb dataset maximum that you are referencing. 

 

Pro and PPU accounts both used shared capacities, and the 10gb-per-table on gatewayrefresh limit applies to all shared capacities. The only way to avoid this limit is with a Premium capacity, as it is not a shared capacity.

 

@albertoserinf I'd be interested to hear what solution you eventually came up with to avoid encountering the same error message. I imagine you were forced to modify your table-size in the database itself.

Hi @mholloway , considering that a Premium capacity license was (and still is) out of budget, we first decided to switch from "import mode" to "direct query" for tables exceeding this limit.

Result was poor performance for end users. 

So we added some aggregatated tables at datawarehouse level but loosing some details requested from analysts.

Final solutions was moving from an on-prem db to parquet files stored in ADLS gen2 as data source, in that way we are able to reduce size of tables. And we do not need a gateway anymore. 

 

Thanks for the reply alberto, very interesting to hear. We also considered DirectQuery (like you, Premium is out of budget) but we had concerns it would be too slow for the end-user -- so it sounds like you've confirmed our fears in that department! Thanks again for the reply.

Helpful resources

Announcements
Las Vegas 2025

Join us at the Microsoft Fabric Community Conference

March 31 - April 2, 2025, in Las Vegas, Nevada. Use code MSCUST for a $150 discount!

Dec Fabric Community Survey

We want your feedback!

Your insights matter. That’s why we created a quick survey to learn about your experience finding answers to technical questions.

ArunFabCon

Microsoft Fabric Community Conference 2025

Arun Ulag shares exciting details about the Microsoft Fabric Conference 2025, which will be held in Las Vegas, NV.

December 2024

A Year in Review - December 2024

Find out what content was popular in the Fabric community during 2024.