Semantic Model with On - Prem Connection & Partitioning

[Anonymous]

Hello everyone!

Currently, I have a model on Analysis Services that I modify using Tabular Editor. My goal now is to replicate this model on Fabric and create a Semantic Model, while keeping the data source on-prem.

I've successfully deployed the model on Fabric using Import Mode, but I'm unsure about what needs to be done with partitioning. Are partitions necessary? How should they be managed? And how can I refresh a specific partition?

Thanks in advance, 
Dimitra

[v-pnaroju-msft]

Thank you @lbendlin and @rohit1991 for your response.

Hi @Anonymous ,

We would like to inquire if the solution offered by @lbendlin and @rohit1991 has resolved your issue. If you have discovered an alternative approach, we encourage you to share it with the community to assist others facing similar challenges.

Should you find the response helpful, please mark it as the accepted solution and add kudos. This recognition benefits other members seeking solutions to related queries.

Thank you.

[rohit1991 · Accepted solution]

Hi @Anonymous ,

Semantic Model with On-Prem Connection & Partitioning in Fabric

Do You Need Partitioning?

 Yes – If handling large data and need faster refresh.
 No – If data is small and refresh speed isn't an issue.

How to Manage Partitions?

• Define partitions in Tabular Editor.
• Use incremental refresh policies in Fabric.

How to Refresh a Specific Partition?

• Use XMLA Endpoints (if supported).
• Automate with Power Automate or REST API.

Best Approach

• Use DirectQuery for real-time data.
• Try Hybrid Tables (Import + DirectQuery) for balance.

[lbendlin]

Same as you did before, with XMLA calls to specify which partition you want to modify in which mode.  Note that dependent partitions may be refreshed as well (especially when you have forgotten to disable Auto Date/Time)