Data Refresh on Service keeps spinning - sometimes errors, sometimes doesn't

[dunn1373]

Hello, we have a data model that suddenly started taking hours to refresh.  We didn't change anything to the model itself.  Some days it works fine and refreshes in ~30 minutes and some days it just spins and spins for 4+ hours.  It will occasionally even be done and still show as spinning.  We have checked our Gateway multiple times and applied updates.  Today is the 8th time in the last 2 weeks that this has happened.  

Does anyone have ideas?!  It's so hard to troubleshoot when we don't have any errors to go off

UPDATE - it did error out but all it said was...

"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."

[v-jtian-msft]

Hi,@dunn1373 I am glad to help you.
Hello,@SaiTejaTalasila.,thanks for your concern about this issue.

Your answer is excellent!
And I would like to share some additional solutions below.

I sympathize with you, it's very frustrating to have a scheduled refresh configured and then have the refresh operation go off and on.
I think that in this case, the possible reason is not in the user's own refresh configuration, because you mentioned that the refresh is able to be executed successfully, which means that there is no problem with the whole refresh process, from the cloud service to the local gateway connection and then to the data source.
Therefore, there may be other reasons why you sometimes fail to refresh.
I think the following reasons are the scope of your consideration
1. server load is too large: if your service is located in the region of the server workload is too large, the current time period needs to execute a large number of users to apply for the refresh program, then this may be the cause of refresh failure
2. The execution of plan refresh needs to apply to enough resources, if you need to refresh the data is too large, sometimes can not apply to enough resources to run the refresh operation, then it may lead to plan refresh failure. In this case, you can try to avoid this situation by reducing unnecessary data in the report.
3. Network latency fluctuations, which are less likely but should be taken into account.
4. For these cases you can get the relevant information by troubleshooting, I recommend that you use the gateway Application, check the Power BI gateway logs to see if there is any reason.
5. If possible, for some often failed to refresh the report, but the configuration is not a problem, you can try to use the way to create a dataflow instead of the original gateway connection, because the dataflow as a cloud “Power query” operation, it is a kind of data integration of data collection, through the use of dataflows as a data source to complete the refresh. By using the dataflow as the data source to complete the refresh, you can bypass setting up the gateway and avoid some gateway performance issues.

URL:
Creating a dataflow - Power BI | Microsoft Learn

I hope my suggestions give you good ideas, if you have any more questions, please clarify in a follow-up reply.

Best Regards,

Carson Jian,

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[SaiTejaTalasila]

Hi @dunn1373 ,

Change your refresh schedule and see whether there is any difference in the refresh.Check the refresh time with personal gateway whether there is any refresh time difference.

Thanks,

Sai Teja 

[dunn1373]

@coneal I submitted a ticket to Microsoft so I can report back if I get something that may help you.  I suggest you do the same if you can

[coneal]

I don't have an answer, but I'm having this issue as well. Latest gateway has been installed. I run the refresh in desktop to pinpoint the exact query causing the issue. Query runs fine in SSMS (10 seconds) gets stuck on 'Evaulating' in desktop and I assume the same in service.