Embedded Capacity with Shared PBI Dataset

[Anonymous]

Hello,

I'm currently working on a large project with a rather complex model. To help speed up development time I was hoping to be able to make edits to a single Power BI dataset while my co-worker works on a report that uses that dataset as its source. Ideally, we would have about five or six reports using this single "golden dataset."

The issue I have run into is that reports using a Power BI data set do not seem to work on our website, where they use embedded capacity. I cannot determine if this is a problem with our API or if shared datasets simply don't work on embedded? Documentation on DirectQuery for PBI datasets says that it is not supported for embedded capacity, but the documentation suggests that this sort of query is only used for composite models. It says that the connection to a PBI dataset uses "Live Connect."

Has anyone used a shared dataset on embedded?

Right now it pulls the error message "404 Not Found: Error while retrieving Embed URL
Not Found: {"error":{"code":"PowerBIEntityNotFound","pbi.error":{"code":"PowerBIEntityNotFound","parameters":{},"details":[],"exceptionCulprit":1}}}"

Barring the ability to use a shared dataset, is there a way to work on a PowerBI report as a shared dataset and then switch it over to its own import dataset? For example, getting rid of the PBI data source and copying and pasting in all the queries from the original dataset?

The shared dataset is really ideal because a lot of the work is in DAX and this would allow us to share measures and calculated tables/columns instead of copying and pasting them into new reports. The Dataflows allow the M side of the report to be shared but this is only half the word and the Dataflows run so incredibly slow they are almost unusable. 


Follow up question: If you have a very large shared dataset but a report only uses visuals from a small part of it, does the size still hurt embedded capacity usage when rendering visuals? It's too bad there isn't an option to truncate a model.

Thanks!

[v-yangliu-msft · Accepted solution]

Hi  @Anonymous ,

The shared dataset is the Live Connection connection mode. According to the research, the Live Connection connection mode cannot be converted to the Import connection mode. The size of the data also affects the use of embedded capacity.

Import models must be fully loaded into memory so they can be queried or refreshed. The Power BI service uses sophisticated algorithms to manage memory usage fairly, but in rare cases, the capacity can get overloaded if there are insufficient resources to meet customers' real-time demands. While it's possible for a capacity to store many import models in persistent storage (up to 100 TB per Premium capacity), not all the models necessarily reside in memory at the same time, otherwise their in-memory dataset size can easily exceed the capacity memory limit. Besides the memory required to load the datasets, additional memory is needed for execution of queries and refresh operations.

https://docs.microsoft.com/en-us/power-bi/enterprise/service-premium-what-is#how-capacities-function

According to the error it shows:

https://docs.microsoft.com/en-us/power-bi/developer/embedded/embedded-troubleshoot#typical-errors-when-embedding-for-power-bi-users

• Wrong Report ID
• Report doesn't exist

This is the related document, you can view this content:

https://community.powerbi.com/t5/Service/PowerBIEntityNotFound-Error/m-p/1218885

https://stackoverflow.com/questions/64109695/getting-powerbi-embed-token-403-forbidden

https://community.powerbi.com/t5/Developer/Paginated-Reports-Embed-getting-error-quot-PowerBIEntityNotFound/m-p/1067484

https://community.powerbi.com/t5/Service/Some-of-the-powerBi-embeded-reports-returns-NotFound-404/m-p/2093342

Best Regards,

Liu Yang

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

[Anonymous]

Thanks for the response. I have figured out what the problem was, it has to do with the way our specific website is built and we have resolved the issue. 

I do still have a question vis-a-vis model size and shared datasets. For an easy example, let's say my 'golden' data set needs to have 20 tables and contains 9 million rows total. However, a given report built using that dataset only uses 5 of the tables, basically, one section of the model. Does being part of a larger model necissarily increase the resources needed to run the report, or is only the data needed to render the visuals called up?

We utilizie a lot of embedded links in our tables/matricies that filter our reports down to a small subset of our data on load. I was hoping this would reduce capacity issues.