power Bi desktop ran out of memory trying to save the data model

[satya_bhotra]

Getting error message when trying to save the data model.

Unable to save the 

[Anonymous]

Your ideas is so great @Shosher 

Hi, @satya_bhotra 

As stated in the error message, it may be due to a memory allocation failure.

This is usually caused by running out of memory on behalf of the hardware to process a given task.
It is recommended that you increase the available memory.

You can also try removing unused columns and filtering data to reduce model size in PowerQuery.

Other related threads：

Solved: Power BI Desktop ran out of memory trying to save ... - Microsoft Fabric Community

Solved: Memory error: Memory Allocation failure. - Microsoft Fabric Community

Solved: Not Enough Memory Space - Microsoft Fabric Community

Solved: Report out of memory - Microsoft Fabric Community

Best Regards

Jianpeng Li

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

[Brunner_BI]

Here is a blog post how to remove all unused columns to optimize your model. @satya_bhotra 

[Shosher]

Lets start with where you are developing, Is it a virtual machine? A remote desktop enviroment? A cheapo company laptop with a grand total of 3gb of ram. 

Whats the dataset size? we working with millions of rows or just a couple thousand. From there we might be able to help allot better!

As a quick side solution you can download Tabular Editor 2 and connect the file directly. From that point you can save the content as TOM file structure. That should save any changes you have so far while you figure out the problem. Dont want you to lose any of that work on what's likely the end of your workday!


Dont forget to set me as the answer if this helped!

[satya_bhotra]

Hi Shosher,

Thank you for looking in this matter.

Its a Citrix desktop (Private XD). Dataset size is 1027 kb. 

its more than millions of rows .

The user is using the import concept to pull table < 5000 rows  and Direct Query to pull > 5000 rows connecting to oracle database.

Regards,

Satya

[Shosher]

Im sorry its tough to say at this point. It could be that the data connector is simply relaying a comment made from your oracle database, without screenshots i really couldnt tell. 

Try using a different device to replicate the issue. Otherwise switch to import mode for now. Hybrid models have never been that stable in practice, especially  when using third party datastorage.