Dataflow and Dataset Refreshed Successfully, but Report Didn’t Update

[euow9302]

I created multiple dataflows connected to different sources (Cloud and SQL), all scheduled to refresh at 6:30 AM. I also have a dataset with the source solely from the dataflow, scheduled to refresh at 7:30 AM. Both the dataflows and the dataset refreshed successfully with no errors and contain the latest data.

However, some visualizations in the report did not reflect the updated data. When I downloaded the report as a PBIX file, it immediately showed the latest data without needing to refresh, which confirms that the dataset itself is up to date. Only after manually triggering a dataset refresh in the Power BI Service did the report display the latest data.

Interestingly, when I duplicated the dataset and report into My workspace with the same schedule, everything updated correctly without any issues.

Why is this happening? FYI, this happened in Pro workspace.

[euow9302]

Hi @v-hjannapu ,

I managed to find the root cause.

It is all because of the calculated column with TODAY() in it of which made the difference between power bi desktop and service.

Thanks for the advice and assistance given. Appreciate it.

[v-hjannapu]

Hi @euow9302,

Good to hear that your issue is resolved and you found the root cause. Thanks for sharing the update as well it will help others too.

If you face any issues in the future, please feel free to reach out here. We are always happy to help.

Regards,
Community Support Team.

[Peter_23]

Hi @euow9302  if you click in update button in the upper right. 

[euow9302]

Have tried and it's not working. 

[Peter_23]

It has a RLS?

[euow9302]

No, no RLS.

[Peter_23]

What's your fabric capacity?

[euow9302]

I'm not in fabric yet. The workspace is in Pro and I'm using dataflow gen1

[Peter_23]

Do you have filters inside the visuals?  is the countrows same online and offline?

[euow9302]

I do have filters on the visuals. What do you mean by countrows? The countrows in the visual are not same online and offline. Online showed yesterday's data but offline showed today's. 

[v-hjannapu]

Hello @euow9302,
Thank you for posting your query in Microsoft Fabric Community Forum. Also, thanks to @Peter_23 , for those inputs on this thread. 

The row count is different in Service and Desktop, it looks like the online report is still showing old data or one of the filters is behaving differently.

You can try this once:
In the Service, select the visual and check the filters on the right side. Remove them for a moment and see if the number changes.

If it’s still the same, please republish the same PBIX to the same workspace. Sometimes the Service keeps old cache, and republishing forces it to load the latest data.

After publishing, open the report again and check if the numbers match. 

Hope this helps if you have any queries we are  happy to assist you further.
Regards,

Community Support Team.

[euow9302]

@v-hjannapu , 

I’ve tried removing the filters, and while the numbers did change, they changed based on old data. When I republished the dataset and report (both in the same file), the latest data appeared correctly. However, after the scheduled refresh ran the next day, the report showed stale data again.

I performed a few tests by adjusting the scheduled refresh times as follows:

1. Dataflow refresh at 6:00 AM, and dataset refresh at 7:00 AM, 7:30 AM, and 8:00 AM

2. Dataflow refresh at 6:30 AM, and dataset refresh at 7:00 AM, 7:30 AM, and 8:00 AM

Only the dataset refresh triggered at 8:00 AM successfully updated the visuals.
Why is this happening?

[v-hjannapu]

Hello @euow9302,

Thanks for sharing the test results. From what you tried, it looks like the dataset refresh is happening too close to the dataflow refresh. Even though the dataflow shows “successful,” sometimes the data is not fully ready when the dataset picks it up. Because of that, the dataset ends up loading the older data, and the visuals in Service also show the old numbers.

The refresh that ran at 8:00 AM worked because it happened after everything in the dataflow was fully completed, so the dataset picked the correct latest data.

For now, please keep a bigger gap between the dataflow and dataset refresh around 45 minutes to 1 hour. This usually avoids the timing clash.

Hope the above provided information help you resolve the issue, if you have any further concerns or queries, please feel free to reach out to us.
Regards,
Community Support Team.

[euow9302]

@v-hjannapu 

I've updated the refresh time of dataflow to 5:00AM and dataset to refresh respectively at 7:00AM, 7:30AM and 8:00AM. Still, only the dataset refresh triggered at 8:00 AM successfully updated the visuals - it seems like 8:00AM is the "magic" time.

I don't think the time gap between the dataflow and dataset is the issue. The dataset is updated when I downloaded the report. It is unsure why the report in service didn't reflect it.

[v-hjannapu]

Hi @euow9302,
Since the dataset has the latest data but the report in Service still shows old values even after all the checks, it looks like something specific to the Service behaviour in this workspace. At this point, I would request you to raise a support ticket with Microsoft. The support team can check the backend logs for this workspace and see why the visuals are not refreshing after the scheduled runs.
Please refer below link on how to raise a contact support or support ticket. 
How to create a Fabric and Power BI Support ticket - Power BI | Microsoft Learn 

Regards,
Community Support Team.

[euow9302]

@v-hjannapu,

Is there any other option other than raise a support ticket? My organisation does not have Microsoft Fabric Subscription or Unified/Premier support plan.

Or is there any other alternatives for me not to use dataflow? My dataset comprise of various sources which involve on premises (sql) and cloud connection (sharepoint and google sheet).

[v-hjannapu]

Hi @euow9302,

I understand your point. Since your organisation doesn’t have a Fabric or Premier support plan, raising a support ticket may not be possible for you.

In this case, the main idea is to avoid the timing dependency between your Dataflow and Dataset refresh. You can try a few simple alternatives:

If possible, move your transformations directly into the dataset (Power Query). Then everything refreshes in one place, and you don’t need the dataflow. or  let all your sources SQL/SharePoint/Google Sheet refresh directly in the dataset and keep a small time gap in the refresh schedule.
If you still want to use Dataflow, make sure the dataset refresh starts only after the Dataflow is fully finished. Without a support plan, this has to be managed manually through timings.

At the moment, these are the practical options available without a support plan.

Hope the above provided information help you resolve the issue, if you have any further concerns or queries, please feel free to reach out to us.
Regards,
Community Support Team.

[v-hjannapu]

Hi @euow9302,
I hope the above details help you fix the issue. If you still have any questions or need more help, feel free to reach out. We are always here to support you.


Regards,
Community Support Team.