Forum Discussion
Power BI Desktop Multiple partitions
Hi all,
I would like to create a power bi report with a table with two partitions :
- Partition 1 based on SQL server 1 and refreshed each 30 minutes
- Partition 2 based on SQL server 2 and refreshed each day
Do you know the easiest way to do it (the creation of the table and the publication).
I tried to create the partition 2 with ssms, it worked but at each publication of the report the partition 2 is deleted.
Any advise ?
Many thanks
Hi GJA ,
The Power BI Deployment Pipeline replicates the entire semantic model (including model metadata, tables, relationships, metrics, partitions, etc.) at push time. This means that any changes you make in the development or test environment (including partition settings) will be deployed along with it when pushed to the next stage. As a result, the deployment pipeline affects the entire semantic model, not just one part.
The Microsoft Fabric deployment pipelines process - Microsoft Fabric | Microsoft Learn
You can use XMLA endpoints in conjunction with Tabular Editor to set up your partitions, here are some procedures to help you!
After a successful connection, you can see the dataset in your workspace. In the Tabular Editor, you can right-click on a dataset and select Partitions, then view, create, or modify it. For example:
You can define different partitions for a table to control how often the data is refreshed.
You can also manually refresh specific partitions to avoid refreshing the entire model.If you still have questions you can check out the document I have prepared for you, I hope it will help you with your doubts!
Hope it helps!
Best regards,
Community Support Team_ Tom ShenIf this post helps then please consider Accept it as the solution to help the other members find it more quickly.
4 Replies
Hi GJA
Since republishing from Power BI Desktop overrides partition settings, use Tabular Editor or SSMS to directly edit the published model on the Power BI Service when changes are needed. Avoid publishing the dataset from Power BI Desktop to prevent partition loss. Instead, work with Deployment Pipelines if additional updates are necessary for the dataset.
Hi Angith_Nair ,
thank you for your advises. What do you mean exatly with deployment pipelines ? A deployment pipeline will affect all the semantic model, not only a part of the semantic model ?
Hi GJA ,
The Power BI Deployment Pipeline replicates the entire semantic model (including model metadata, tables, relationships, metrics, partitions, etc.) at push time. This means that any changes you make in the development or test environment (including partition settings) will be deployed along with it when pushed to the next stage. As a result, the deployment pipeline affects the entire semantic model, not just one part.
The Microsoft Fabric deployment pipelines process - Microsoft Fabric | Microsoft Learn
You can use XMLA endpoints in conjunction with Tabular Editor to set up your partitions, here are some procedures to help you!
After a successful connection, you can see the dataset in your workspace. In the Tabular Editor, you can right-click on a dataset and select Partitions, then view, create, or modify it. For example:
You can define different partitions for a table to control how often the data is refreshed.
You can also manually refresh specific partitions to avoid refreshing the entire model.If you still have questions you can check out the document I have prepared for you, I hope it will help you with your doubts!
Hope it helps!
Best regards,
Community Support Team_ Tom ShenIf this post helps then please consider Accept it as the solution to help the other members find it more quickly.
Hi Anonymous
It looks like a good solution. I'm going to try it.
Thank you
