Incremental Refresh Missing Data

[cchilton]

Hello! 

https://learn.microsoft.com/en-us/power-bi/connect-data/incremental-refresh-configure

I followed the Power BI docs on incremental refresh, which recommend using a small RangeStart/RangeEnd in Desktop, since those values are supposedly overridden in the Service. But this caused a major issue in our embedded report: data outside that small range (like from 5 years ago) was missing entirely—even after publishing.

Turns out:

• The model published to the Service only includes what was loaded in Desktop.

• The refresh policy only kicks in after the first full refresh in the Service.

• So if you publish before doing a full refresh and used a small range, older data simply doesn't exist in the model, and visuals pulling from it will be blank.

Fix: I temporarily set RangeStart to 1900 and RangeEnd to 3000 in Desktop, reloaded all data, and republished. Then incremental refresh worked as expected after the first scheduled refresh. Would be great if the docs made this clearer for embedded scenarios.

What am I missing? Am I doing this correctly? Would love to hear how others are handling this, especially in embedded scenarios where we can’t afford to have data missing after publishing.

[lbendlin · Accepted solution]

You are not missing anything.  Here are your options

- do your republishing during low traffic periods so fewer users are impacted by the temporary loss of data

- abstain from changing your report design

- use ALM Toolkit to synchronize meta data, and use the Enhanced Refresh API to selectively refresh individual partitions

[v-venuppu]

Hi @cchilton ,

I hope this information is helpful. Please let me know if you have any further questions or if you'd like to discuss this further. If this answers your question, please accept it as a solution and give it a 'Kudos' so other community members with similar problems can find a solution faster.

Thank you.

[v-venuppu]

Hi @cchilton ,

I wanted to check if you had the opportunity to review the information provided. Please feel free to contact us if you have any further questions. If my response has addressed your query, please accept it as a solution and give a 'Kudos' so other members can easily find it.

Thank you.

[v-venuppu]

Hi @cchilton ,

May I ask if you have resolved this issue? If so, please mark the helpful reply and accept it as the solution. This will be helpful for other community members who have similar problems to solve it faster.

Thank you.

[ribisht17 · Accepted solution]

HELLO @cchilton 

Please check Solved: Incremental Refresh | RangeStart vs RangeEnd Param... - Microsoft Fabric Community 

Regards,

Ritesh

Community Champion

Please mark the answer if helpful so that it can help others

Dance-Sing with Data Projects - YouTube

[lbendlin · Accepted solution]

You are not missing anything.  Here are your options

- do your republishing during low traffic periods so fewer users are impacted by the temporary loss of data

- abstain from changing your report design

- use ALM Toolkit to synchronize meta data, and use the Enhanced Refresh API to selectively refresh individual partitions