Forum Discussion
insufficient memory to complete this operation
Hello everyone,
I get an error on power bi Desktop when updating data: Failed to save changes on server. Error returned: "There is insufficient memory to complete this operation. Try again later at a time when there may be more memory available". even though I haven't made any changes on my desktop.
The pbix is 34 Mb. The data is a maximum of 8000 lines returned by the Dynamics 365 oAuth2.0 api.
This has been happening since mid-September.
Do you have a solution to find out what's causing the problem?
I don't have any particular DAX calculation. Everything is done on Power M.
Thank you very much.
Hello,
I found the solution to the performance problem.
First, My power bi file called an external dataset in a dataset that takes more than 5 hours to refresh. This was causing a memory saturation problem. The solution was to separate the visual part from the dataset and call the dataset that was causing the memory saturation problem from the visual dataset.
Secondly, on the dataset from my Dynamics 365 CRM source. I created tables so as not to use the Table.ExpandRecordColumn function on large tables, but to use the Table.NestedJoin and then Table.ExpandTableColumn functions to retrieve the necessary information. This limits the number of calls to the webservice and avoids repeated attempts to connect to the webservice. The loading time was multiplied by the multiple connections to the Dynamics 365 webservice.
Thirdly, changes to my desktop file, which was currently being modified, were not being taken into account. So I opened the pbix zip file, deleted the UnappliedChanges file and reapplied the M code to each of my tables.
The result is a drop in loading time from over 7 hours to 48 minutes on powerbi desktop and powerbi services.
The dataset file still contains the 130 tables required.
Thank you for your help
4 Replies
Hello Anonymous ,
check this solution https://community.fabric.microsoft.com/t5/Desktop/Insufficient-memory/m-p/1381881
If I answered your question, please mark my post as solution, Appreciate your Kudos 👍
Follow me on Linkedin
Vote for my Community Mobile App Idea 💡Hello,
I tested the solution but unfortunately it doesn't solve my problem. I still get the same message "There is insufficient memory to complete this operation. Try again later at a time when there may be more memory available." on Dax Studio, the same as on power bi Desktop. I've tried to increase the memory of my workstation from 8GB to 16GB. I'm on premium capacity per user.
If you have any other tips, I'd love to hear from you.Hello Anonymous ,
I would suggest to reach to microsoft support directly through this link https://support.powerbi.com
If I answered your question, please mark my post as solution, Appreciate your Kudos 👍
Follow me on Linkedin
Vote for my Community Mobile App Idea 💡
Hello,
I found the solution to the performance problem.
First, My power bi file called an external dataset in a dataset that takes more than 5 hours to refresh. This was causing a memory saturation problem. The solution was to separate the visual part from the dataset and call the dataset that was causing the memory saturation problem from the visual dataset.
Secondly, on the dataset from my Dynamics 365 CRM source. I created tables so as not to use the Table.ExpandRecordColumn function on large tables, but to use the Table.NestedJoin and then Table.ExpandTableColumn functions to retrieve the necessary information. This limits the number of calls to the webservice and avoids repeated attempts to connect to the webservice. The loading time was multiplied by the multiple connections to the Dynamics 365 webservice.
Thirdly, changes to my desktop file, which was currently being modified, were not being taken into account. So I opened the pbix zip file, deleted the UnappliedChanges file and reapplied the M code to each of my tables.
The result is a drop in loading time from over 7 hours to 48 minutes on powerbi desktop and powerbi services.
The dataset file still contains the 130 tables required.
Thank you for your help
