Automated refresh 4 tables from semantic model

[ahmedshalabyy12]

Dears , 

i know how to schedule refresh for the whole model but i need to schedule refresh for only 4 tables from the semantic model
to be every 30mins  so how i can do that

Thank you

[MasonMA · Accepted solution]

@ahmedshalabyy12 , Hi, probably no, but one of the approaches I tried in the past is split the frequently refreshed tables into a separate dataset and use composite models or DirectQuery for Power BI datasets to combine them in your reports.

For example, Dataset1 has 4 frequently refreshed tables, then refresh every 30 mins. Dataset2 has other slow-changing tables, then refresh daily or weekly. 

This gives fine-grained refresh control at the expense of model complexity.

[v-ssriganesh]

Hi @ahmedshalabyy12,

Thank you for posting your query in the Microsoft Fabric Community Forum, and thanks to @lbendlin & @MasonMA for sharing valuable insights.

Could you please confirm if your query has been resolved by the provided solution? If so, please mark it as the solution. This will help other community members solve similar problems faster.

Thank you.

[ahmedshalabyy12]

Yes solved both considered as solution thank you 

[lbendlin · Accepted solution]

Easy. Both in Power Automate and in data pipelines you can specify the refresh at model, table, and even partition level (for incremental refresh scenarios).

In Power Automate you need to write your own connector for the enhanced refresh API, in Fabric you get that out of the box.

Enhanced refresh with the Power BI REST API - Power BI | Microsoft Learn

[MasonMA · Accepted solution]

@ahmedshalabyy12 , Hi, probably no, but one of the approaches I tried in the past is split the frequently refreshed tables into a separate dataset and use composite models or DirectQuery for Power BI datasets to combine them in your reports.

For example, Dataset1 has 4 frequently refreshed tables, then refresh every 30 mins. Dataset2 has other slow-changing tables, then refresh daily or weekly. 

This gives fine-grained refresh control at the expense of model complexity.