Forum Discussion

jgarciabu's avatar
jgarciabu
Advocate I
9 years ago
Solved

visual has exceeded available resources

Hi All,   I've been using PowerBI for the better part of a year now. For the last month and a half or a little more, we've had annoying failures to refresh for 6 of our 14 tile visualizations. It's...
Need Help
  • v-ljerr-msft's avatar
    v-ljerr-msft
    9 years ago

    Hi jgarciabu,

     

    Yes, the issue occurs when a visual has attempted to query too much data for the server to complete the result with the available resources.

     

    As suggested in the error, you may need to try filtering the visual to reduce the amount of data in the result currently.:smileyhappy:

     

    Regards

CahabaData's avatar
CahabaData
Memorable Member
9 years ago

as I understand the reply from Microsoft - it is the size of the record set.  Not the dax or visuals per se.  So i.e.  instead of trying to pull in X years of data - segment that / filter to a smaller data set.

 

but what is not said is how much is too much.... that would be worthy of note.....

  • malagari's avatar
    malagari
    Continued Contributor
    9 years ago

    There are definitely some limitations in regards to in-memory analysis as well.  I imagine that Power BI is running Azure Analysis Services behind the scenes and that a certain amount is allocated to each user.

     

    My tabular dataset isn't huge.  I've seen this when I'm doing a significant number of date filters in a DAX measurement to create a date-dynamic measurement for historical trends.  It takes a good chunk of my RAM locally when running, but completely borks when it's published.

     

    I guess the next step would be to set up my own SSAS for my tabular model, rather than relying on Power BI's.

    • CahabaData's avatar
      CahabaData
      Memorable Member
      9 years ago

      interesting to hear.  in your comparison - how much RAM locally is it taking a good chunk of??  are you up in the 32G area on a server or more?

      • malagari's avatar
        malagari
        Continued Contributor
        9 years ago

        I'm able to run some of the measurements without issues on my 12GB RAM Windows VM, but it doesn't cooperate when I have multiple card visuals with different measurements on the same report page.  On my desktop, I have 32GB of RAM and I'm able to render everything properly.

         

        The "big" DAX calculation can take upwards of 18GB of RAM when making the date comparisons.  It really only gets bad on my desktop when I have more than one visual doing a fairly large calculation against a table of ~40K rows.

         

        It's at the point where I'm investigating if OLAP Cubes would help improve the performance of my historical data (again, offloading the strenuous calculations to an SSAS instance).