Forum Discussion

jgores77's avatar
jgores77
Regular Visitor
8 years ago

Possible memory leak?

 

Hello,

 

I'm running a DirectQuery to a database using 4 tables from the same database.  When I try to add latitude and longitude columns from one of the tables, it adds them fine, but creates an average of the values.  Then when I tell Power BI "Don't Summarize" these columns it cranks for a little while and does it, but then the memory usage climbs until all memory is used and remains there for a long time.  Very often I simply have to close Power BI and wait for the memory usage to drop back down and then start over.  I have experienced this with other columns and other tables in this database, every time it is the msmdsrv.exe process that is eating up the memory.  (See attached screen shot.)

 

What is happening?  How do I fix this?

 

Thanks!

6 Replies

  • Anonymous's avatar
    Anonymous
    Not applicable

    HI jgores77,

     

    I think your table contains huge amount of records, right?

     

    When you cancel summary method, power bi will expand summarized records and re-analysis and calculation formulas on them.(these operation will cost huge amount of memory if your table contains large amount records, complex measure calculation, nested loop functions)

     

    BTW, power bi will generate analysis instance to store your data when data model generated. So 'msmdsrv.exe' will cost lot of memories when power bi calculation these process.

     

    Reference:

    What is msmdsrv.exe?

     

    Spoiler
    The process known as Analysis server or Microsoft SQL Server Analysis Services or SQL Server Analysis Services (MSSQLSERVER) or OLAP server or Microsoft SQL Server (version 2005 Analysis Services) 

    belongs to software Microsoft SQL Server Analysis Services or Microsoft SQL Server(version 2005, 2005 Books Online, 2005 Tools, 2005 Backward, 2005 Analysis Services) or Microsoft SQL Server Native or Microsoft SQL Server OLAP Services 

     

    Regards,

    Xiaoxin Sheng

    • jgores77's avatar
      jgores77
      Regular Visitor

      Hi Anonymous/Xiaoxin,

       

      I thought this might be the case as well, so I limited the data to a single year using the Page Level Filter (reducing my rows of data from 600K to about 14K) and deleted all calculations.  Power BI still does the same thing.

       

      Any other thoughts?

      • Anonymous's avatar
        Anonymous
        Not applicable

        HI jgores77,

         

        If you only drag that column to visual and do same operate it, do this issue appears?

        If your create new report or test with other device, did these operations spend less memory?

         

        Regards,

        Xiaoxin Sheng