Forum Discussion

david-hepstar's avatar
david-hepstar
Regular Visitor
2 years ago
Solved

46GB PPU Semantic Model: "Memory error: maximum allowable memory allocation for your tier"

I have one large Semantic Model with data refreshed incrementally from Google BigQuery. Partitions are by month for the last 24 months, and by default the report refreshes the last 14 days (usually s...
Need Help
  • david-hepstar's avatar
    david-hepstar
    2 years ago

    I've finally received a different error message, which I assume is closely related to the ones I was receiving earlier. This makes me think that there may be some user settings I could adjust to get around this, though I haven't confirmed this.

    Failed to save modifications to the server. Error returned: 'Resource Governing: This operation was canceled because there wasn't enough memory to finish running it. Either reduce the memory footprint of your dataset by doing things such as limiting the amount of imported data, or if using Power BI Premium, increase the memory of the Premium capacity where this dataset is hosted. More details: consumed memory 53639 MB, memory limit 49315 MB, database size before command execution 53084 MB. See https://go.microsoft.com/fwlink/?linkid=2159753 to learn more.

Technical Details:
RootActivityId: 344ed8dd-99a4-4539-acbb-392119ea557e
Date (UTC): 3/23/2024 11:06:16 AM

     
    In the end I've avoided this issue entirely by redesigning my data model to incorporate a second fact table at a different grain. This enabled me to drop an expensive ID column from my original fact table, thereby cutting my memory usage by ~85%.

collinq's avatar
collinq
Super User
2 years ago

Hi david-hepstar ,

 

According to this article, Power BI Premium Per User - Power BI | Microsoft Learn, the 100 GB limit for PPU may also include some things reserved for other operations.  Also, since you stated it does not always error, I am wondering if anything else anywhere with your PPU license is running?  It wouldn't take much for another refresh of a different report to be running at the same time to consume all the capacity.  (according to the documentation "A Premium capacity might throttle data refresh operations when too many semantic models are being processed concurrently.")

 

Also, according to this article,  What is Power BI Premium? - Power BI | Microsoft Learn for non-Premium capacities, the amount of memory required for semantic model is doubled.

 

Also, it is possible that within the queries themselves you have created some sort of calculation or measure (or a circular reference) that could be running amok.  Has anything changed in the query side since the last successful run?

 

Another item to check is your gateway.  Is it being overwhelmed?

 

 

  • david-hepstar's avatar
    david-hepstar
    Regular Visitor
    2 years ago

    Thanks for this feedback collinq. Still struggling to pin this down, but you've brought up some helpful ideas, which I'm in the process of investigating. A couple of comments:

     

    Also, since you stated it does not always error, I am wondering if anything else anywhere with your PPU license is running?

     

    Very little else happening in this Workspace. I have another non-PPU workspace with some dataflows. I don't think they're big, and they aren't refreshing. I'm not sure how to check their size either. I think as a next step I may try and delete these, just in case. It's also not clear to me if these dataflows contribute to my 100GB memory limit, especially since they're in a non-PPU workspace.

     

     

    (according to the documentation "A Premium capacity might throttle data refresh operations when too many semantic models are being processed concurrently.")

     

    Thanks for pointing this out, I hadn't thought of this. I do have another small (32 MB) semantic model in the workspace, but it is only refreshed manually and very infrequently. I have another small data model in another workspace that refreshes hourly. It seems like that shouldn't conflict, but I suppose it might. 

     

     

    Also, according to this article,  What is Power BI Premium? - Power BI | Microsoft Learn for non-Premium capacities, the amount of memory required for semantic model is doubled.

     

    I don't think this applies here since I am working within Premium Per User and only refreshing part of the model. If I've done my sums right, the data model takes up 46 GB. The incremental refresh would take up another ~6 GB. And I'm wondering where the other 48 GB has gone to? I can imagine that some heavy querying might cause issues from time to time, but it seems unlikely given I'm the primary user.

     

    Also, it is possible that within the queries themselves you have created some sort of calculation or measure (or a circular reference) that could be running amok.  Has anything changed in the query side since the last successful run?

    Another good idea. No calculated columns in the model at all. There are some measures I know are slow, but not significantly slow, and I'm not sure these would have much impact if I'm not generally using them during the model refresh.

     

     

    Another item to check is your gateway.  Is it being overwhelmed?

     

    Helpful suggestion, thanks! I'm using a cloud connection to Google BigQuery. If I'm understand correctly, this does not involve any gateway, so there's not much for me to troubleshoot or configure here. 

     

    • collinq's avatar
      collinq
      Super User
      2 years ago

      Hey david-hepstar ,

       

      You are right - you are not using a gateway if that is your only connection.

       

      It does sound like you don't have much concurreny either.

       

      I think it might be time to open a ticket and have Microsoft help you track down the issue.  let us know the result if you go that direction.