Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Refreshes failing in Power BI service - "internal service error"

We are experiencing a rash of refreshes failing.  This started Friday (8/13). 

It seems to maybe be related to us upgrading to the latest desktop version.  The desktop Power Query editor seems to be really slow as well and datasets that we haven't opened and republished from the new desktop version seem to be working as expected.
We are using Amazon Redshift as our source database.

 

Here are 2 examples of the failures that are happening:

"CASE"

Something went wrong

The last refresh attempt failed because of an internal service error. This is usually a transient issue. If you try again later and still see this message, contact support.
Please try again later or contact support. If you contact support, please provide these details.
Cluster URIWABI-US-NORTH-CENTRAL-redirect.analysis.windows.net
Activity ID486a8951-de71-4b81-a60c-f13a740e5ab9
Request IDd3217abf-8bdb-4263-9ce7-9f3ea1371640
Time2021-08-16 15:43:28Z

 

"SUBSCRIPTION"

Something went wrong

The last refresh attempt failed because of an internal service error. This is usually a transient issue. If you try again later and still see this message, contact support.
Please try again later or contact support. If you contact support, please provide these details.
Cluster URIWABI-US-NORTH-CENTRAL-redirect.analysis.windows.net
Activity IDbe29f429-b2d6-49df-b73e-6a1c626e2b6c
Request ID98661157-c15b-3e56-5458-e95078d3b5c7
Time2021-08-16 16:11:37Z

 

 

 

Status: New
Comments
v-yingjl
Community Support

Hi @ScottBeal ,

This issue has been confirmed as a temporary issue in Power BI. You can refresh the report in some other time or use previous version of Power BI Desktop to republish the report to refresh to check it again.

 

If it still has the same issue many times, you can use Fiddler to collect some detail error message when the refresh has failed.

 

Best Regards,
Community Support Team _ Yingjie Li

 

 

gizzyuk74
New Member

same here

 

Activity ID: 4b983940-edce-497d-9650-a8f7617a2e89
Request ID: 93379ac4-94da-2b5d-21e1-daa30363486c
Correlation ID: a3359c4a-4fb3-57d1-febf-65c018a845b8
Status code: 0
Time: Tue Aug 17 2021 14:35:16 GMT+0200 (Central European Summer Time)
Service version: 13.0.16577.55
Client version: 2108.1.07315-train
Cluster URI: https://wabi-west-us-b-primary-redirect.analysis.windows.net/

pjefferi
New Member

@@v-yingjl This isn't a solution.  I tried an earlier copy and received an error message saying that my PBI file was incompatible with the earlier version of PBI.  When will this "temporary" issue be fixed?  I encountered the same issue on two different machines.  Clearly this issue occured with the 8/16/21 update.

jrogersonidea11
New Member

I'm experiencing this since yesterday also for one particular report. Keen to hear when this will be resolved.

 

MicrosoftTeams-image.png

Asum123
Regular Visitor

Hi,

I have a same problem from Mon 16/08/2021 after updating Power BI Desktop to Aug 2021. That day, regular morning refresh worked yust fine and after I updated my Desktop version refreshes just started to produce that error (they are taking a lot of time to get that feedback, around 6,5h per one try).

I tried what @ v-yingjl@ proposed, but it did not worked and had a same error after republishing the model in old version of PBI Desktop.

Now I have to daily manually refresh desktop first (I have no problem there) and replace the model on service to get a refresh done for the end users. This is a lot of manual work and on other side there is a lot of end users for which this is operational tool and it has to be up to date…

Other reports are refreshing just fine.

Error massage:

The last refresh attempt failed because of an internal service error. This is usually a transient issue. If you try again later and still see this message, contact support.

 

Asum123_0-1629358732342.png

 

v-yingjl
Community Support

Still could not reproduce this issue.

Would suggest that create a support ticket here for further if this issue is urgent.

Support Ticket.gif

 

Best Regards,
Community Support Team _ Yingjie Li

 

ScottBeal
Advocate I

@Asum123 We are doing the same thing...refreshing locally and then publishing to the service, for the affected datasets.

@v-yingjl We have opened a support issue.  I was on with the Support Engineer for a while to collect some logs.  A bug ticket has been created for the product team.

 

As I get feedback, I will update this issue.

I appreciate everyones input.  Hopefully they will ge tthis addressed soon.

sarkuma
Employee

Hi Team,
I am facing the same refresh issue since 17/08/21.

Cluster URI: DF-MSIT-SCUS-redirect.analysis.windows.net
Activity ID: 160b4856-0e0d-4872-9e27-1a03c64cf537
Request ID: ed1320db-669a-fde3-27e5-a8882b700431
Time: 2021-08-22 18:08:30Z

 

Keen to hear when this will be resolved.

 

 

 

ysanabria
Frequent Visitor

Same issue here

 

An error happened while reading data from the provider: 'Could not load file or assembly 'System.Runtime.CompilerServices.Unsafe, Version=4.0.4.1, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a' or one of its dependencies. The located assembly's manifest definition does not match the assembly reference. (Exception from HRESULT: 0x80131040)'
ScottBeal
Advocate I

Our issue was narrowed down to a dashboard having a tile related to this dataset on it.
It seems like datasets will encounter errors and not refresh if changes are made that "break" the tile on an existing dashboard.
We are not sure why the dashboard doesn't just break as does a report when the underlying structure changes.
Another issue us that the dashboard seems to exist in a user's "personal" workspace.  Even as an administrator we cannot "fix" (or remove) the dashbaord in order to enable refreshing to continue.  Nor can we even see what workspcae the dashboard is in, in orde rto reach out to the appropriate user(s).

So this is still an issue for us.   We have been manually refreshing the affected datasets daily, locally, and then and publishing to the service.  This makes the current data available.  (just can't reresh in service)