Forum Discussion
Report visuals showing incomplete data while semantic model refresh (incremental ref) is in progress
Hello All
I have a semantic model with Incremental Refresh set-up for last 15 days.
While my report is under scheduled refresh, if an user looks at the report he could only see partial data on the report tables (visual) however when the refresh completes, user is able to view the complete data.
Could you please explain me the reason for the same, so that i can communicate the same to the users?
Thanks in advance.
Hi roysampad91,
Thank you for reaching out to Microsoft Fabric Community.
Thank you Natarajan_M and cengizhanarslan for the prompt response.
On shared capacity, the semantic model refresh and user queries use the same dataset instance. Since there is no query scale-out or read-only replica in shared capacity, some inconsistencies during refresh are expected platform behaviour, not a configuration issue.
- The best possible way is to use Premium or Fabric capacity or schedule refresh during low usage hours.
Thanks and regards,
Anjan Kumar Chippa
8 Replies
Hi roysampad91 ,
PBI by default will isolate the refresh process in the sense it takes a copy of the dataset and performs refresh while the main copy will be serving the users post refresh the updated copy will replace the old .
During this perod you should be seeing the data as per the previous day refresh . check your refresh configuration . it will smoothly transaction the data partitions without out notice .
check query scaleout feature :
https://learn.microsoft.com/en-us/fabric/enterprise/powerbi/service-premium-scale-outThanks but i am on a shared capacity and users complain that for a split second or a couple of mins they could still see that the data is inconsistent. Is there any other workaround?
what Natarajan_M describes is actually a pretty simple workaround in your situation and as far as I know it is the best option in use. As soon as the refresh process completed, users could see refreshed data, while the model is in refresh process it would not affect the usage.
I suspect this can be done for a non-premium capacity? If no, what's the alternative to address the concern of stale data during semantic model refresh is in progress?
Could you pls guide on any solution to this when your workspace is not on a premium capacity?
Hi roysampad91,
Thank you for reaching out to Microsoft Fabric Community.
Thank you Natarajan_M and cengizhanarslan for the prompt response.
On shared capacity, the semantic model refresh and user queries use the same dataset instance. Since there is no query scale-out or read-only replica in shared capacity, some inconsistencies during refresh are expected platform behaviour, not a configuration issue.
- The best possible way is to use Premium or Fabric capacity or schedule refresh during low usage hours.
Thanks and regards,
Anjan Kumar Chippa
Hi roysampad91,
As we haven’t heard back from you, we wanted to kindly follow up to check if the solution I have provided for the issue worked? or let us know if you need any further assistance.
Thanks and regards,
Anjan Kumar Chippa
Hi @roysampad91,
We wanted to kindly follow up to check if the solution I have provided for the issue worked? or let us know if you need any further assistance.
Thanks and regards,
Anjan Kumar Chippa
