refresh error in service not in desktop

[bsheffer]

Suddenly getting a uncompressed table is over 10g error on the service.  Not getting the any error in the desktop.  I can refresh the service after I do the update.

the only think I can think would be different is I have 32G memory on my workstation but I don't know what you get with a pro license on the service.

Any ideas of why this is happening?

[arvindsingh802]

@bsheffer - This is a known issue with Data Refresh. (Data reduction techniques for Import modeling - Power BI | Microsoft Docs)

To permanently get rid of the issue, the only option is to move to Premium capacity. You can also try below methods that will be helpful.

• Only Extract columns from source that are requried. Reduce the Rows also if possible.
• In case you are using Database, change the data type of Columns to lower memory ones.
• Use LookUp tables and In Facts use LookUp keys instead of complete strings.
• Move your calculated columns from PQ to DAX as DAX columns are not included in 10GB limit. 
• If source permits, change from Import to Direct Query

This issue is because of non premium workapce only

Step Beyond the 10GB Limitation of Power BI - RADACAD

If this post helps, then please consider Accept it as the solution and give it a thumbs up

[bsheffer]

I'm trying to import a date and powerbi keeps bringing it in as datetime.  I assume datetime uses more space that date.  I'm already using cast(activity_date as date) in the tsql import statement.  Is there a way to force powerbi import to use date instead of datetime?

[v-kkf-msft]

Hi @bsheffer ,

Try to change the type of the date column to "date" in the Advanced Editor in Power Query Editor, like this:

If the problem is still not resolved, please provide detailed error information or the expected result you expect. Let me know immediately, looking forward to your reply.

Best Regards,
Winniz

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

[bsheffer]

Please confirm:  I can refresh in power bi desktop and publish to the pro workspace (and my dashboards that use this datasource will continue to work correctly) but power bi service will not refresh the data source correctly because it is a pro space?

[arvindsingh802]

Yes, In Power BI Desktop (freeware) there is no limitations, it depend on system confirguration where it is being refreshed - It also not require Power BI Pro, you can sign in but not mandatory to refresh data

All the limitations applied as mentioned in Pricing & Product Comparison | Microsoft Power BI when you are publishing to the Web or refreshing your data on web

If this post helps, then please consider Accept it as the solution and give it a thumbs up

[bsheffer]

arvindsigh802, please respond to my question and not just repost your initial response.

[arvindsingh802]

I would need those infirmation to help you beeter.

[bsheffer]

I've posted a screen shot of the group storage settings

[arvindsingh802]

Can you confirm worksapce type and space of worksapce 

[bsheffer]

it is a power bi pro space and we aren't using anything near the maximum of it.

[bsheffer]

Sorry that didn't help at all.  I can publish to the service from desktop and it works fine so it cannot be a PPU issue or are you saying that this is only an issue if refreshed on the service?

[arvindsingh802]

Reason 1 - You are using Power BI Pro, not PPU/Premium workspace. refer Pricing & Product Comparison | Microsoft Power BI

Reason 2 - If you are using PPU/Premium and not using incremental refresh

   Implemet incremental reresh then you can have your dataset beyon 10 GB in PPU/Premium

If this post helps, then please consider Accept it as the solution and give it a thumbs up

[bsheffer · Accepted solution]

The issue is that one table uncompressed is over the size limit for Pro.  I noticed that the dates were bringing in as datetime even if I stipulated in the source sql that it was date.  Changing from datetime to date after bringing it in won't solve the initial load issue.  Reducing the number of rows did eliminate the error but didn't solve the problem because I don't want to eliminate rows.  My boss hasn't authorized splitting the columns into two tables to reduce the total size.  Therefore as suggested above, the best solution based on my constraints is to move everyone over to the PPU license, essentially doubling the total cost to company for PowerBI.