March 31 - April 2, 2025, in Las Vegas, Nevada. Use code MSCUST for a $150 discount! Early bird discount ends December 31.
Register NowBe one of the first to start using Fabric Databases. View on-demand sessions with database experts and the Microsoft product team to learn just how easy it is to get started. Watch now
Hi,
I have several BI reports in different workspaces that have started to fail during the refresh (scheduled or on demand) since some days ago, with error "There's not enough memory to complete this operation. Please try again later when there may be more memory available."
Some of the BIs that are failing, have 100mb size (not much) and takes 2min to refresh in local.. but when i publish again and try to refresh "on demand" it takes 1h and then fails with similar error. I dont have any problems with the size of the workspace , etc.
Is is possible that there are something wrong or bug in the service that are producing this? it is strange that my BIs that are cofnigured months ago.. started to fail now.
Solved! Go to Solution.
it seems that i dont have the issue since this weekend
We faced multiple dataset refresh failures today. Some of these are small datasets and they have been refreshing regularly without any issues.
Could it be because these are all on shared capacity and compute and memory resources at some occassions fall short? (cases when large number of datasets are refreshing at same times). If there was a easy way to track then this could be avoided by shifting few dataset refreshes by half or one hour. There are many workspaces and many dataset owners. I am not admin so not sure how it seems on admin portal.
i dont have any issues today, but i have some workspaces with 20-25 datasets and i dont have any issue. I think that the only limitation is the capacity per workspace.
In addition, depends of where are you connecting, if you are connecting to a source (example SQL) and the 20 datasets are connecting to the same SQL at same moment.. and other people too.. maybe the SQL are having bad performance due to that.
it seems that i dont have the issue since this weekend
So far Microsoft has acknowledged issues with dataflow refreshes but not dataset refreshes. It's been pretty bad for weeks now, and feel very random.
I feel the whole thing is way oversubscribed/under-resourced, especially Pro but even PPU has random stalling refreshes. To mitigate this, I'm splitting ETL out of datasets and into several dataflows to limit having a single point of failure, though that's not always a practical solution depending on your data sources and data model.
Same issue here, started Wednesday morning for my organization.
Since early this week the refresh of dataset on Powerbi Service takes either more than an hour (usually 5min) or get canceled because the refresh took too much time.
The dataflows refresh are running fine!
And the refresh from PBI Dekstop doesn't have any problem too!
I'm experiencing the same issue here to refresh dataset in PBI Service, location : France
since when? i destect since last weekend.
Even we are facing the exact same issue , it does feel more like this is a bug rather than some data issue
Yes it seems something in the service... becasue in local refresh correctly en few minutes but in service fails... and in service fails during the mornings (spain) but not during the afternoon.. seems strange.
i will wait if someone of microsoft can tell something about this
March 31 - April 2, 2025, in Las Vegas, Nevada. Use code MSCUST for a $150 discount!
User | Count |
---|---|
36 | |
27 | |
19 | |
11 | |
8 |
User | Count |
---|---|
55 | |
43 | |
25 | |
13 | |
12 |