Forum Discussion

bimchahaha's avatar
bimchahaha
Microsoft Employee
5 years ago
Solved

Power BI dataset size limit

So my report (.pbix) has 10 Datasets.... so if I am a user of PowerBI pro. so that report max size is 1GB ???? No matter how many datasets are there in the reports ??? And using import mode ?

  • Anonymous's avatar
    Anonymous
    5 years ago

    Hi,bimchahaha 

     

    I am not very agree with your opinion that your report (.pbix) has 10 datasets.

    One Pbix file corresponds to one dataset. There is a 1 GB limit per dataset that is imported into Power BI. If you have chosen to keep the Excel experience, instead of importing the data, the limit is 250 MB for the dataset.

     

    Please notice that if we say max dataset size is 1 GB - it means the dataset of a single pbix file you are going to publish must be smaller than 1 GB. when you publish a pbix file over 1GB to Power BI Service, you will get limitation error message.

    For more information, please refer to this: https://docs.microsoft.com/en-us/power-bi/connect-data/service-get-data

     

    Depending on the SKU, Power BI Premium supports uploading Power BI Desktop (.pbix) model files up to a maximum of 10 GB in size. You should have at least a P1 or an A4 SKU for any datasets larger than 1 GB.

    For more information, please refer to this: https://docs.microsoft.com/en-us/power-bi/admin/service-premium-what-is#large-datasets

     

     

    Hope it helps.

     

    Best Regards,

    Caitlyn Yan

     

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

3 Replies

  • Anonymous's avatar
    Anonymous
    Not applicable

    Hi,bimchahaha 

     

    I am not very agree with your opinion that your report (.pbix) has 10 datasets.

    One Pbix file corresponds to one dataset. There is a 1 GB limit per dataset that is imported into Power BI. If you have chosen to keep the Excel experience, instead of importing the data, the limit is 250 MB for the dataset.

     

    Please notice that if we say max dataset size is 1 GB - it means the dataset of a single pbix file you are going to publish must be smaller than 1 GB. when you publish a pbix file over 1GB to Power BI Service, you will get limitation error message.

    For more information, please refer to this: https://docs.microsoft.com/en-us/power-bi/connect-data/service-get-data

     

    Depending on the SKU, Power BI Premium supports uploading Power BI Desktop (.pbix) model files up to a maximum of 10 GB in size. You should have at least a P1 or an A4 SKU for any datasets larger than 1 GB.

    For more information, please refer to this: https://docs.microsoft.com/en-us/power-bi/admin/service-premium-what-is#large-datasets

     

     

    Hope it helps.

     

    Best Regards,

    Caitlyn Yan

     

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

  • exactly, each dataset cant be bigger them 1gb and each workspace cant get more tham 10gb of total space (these its multiplied by each existing license of power bi pro but not to tha workspace total but the amount of possible workspaces. )

    using import mode its the one option that consumes more space but load the least on your company servers if they are the source and gateway location. 

  • Adam_Summers_ms's avatar
    Adam_Summers_ms
    Microsoft Employee

    Hey bimchahaha ,

     

    regardless of how many different queries you are importing, 1G is actually a _lot_ of memory if you use it wisely to build a model. The key is not to assume that you just import a table from a source, and do no ETL. For example, if you are importing datetimes, and you only really need the date, then only store the date (and if you need the time, often keep that in a separate field) -- this reduces the cardinality of the column, and saves a LOT of memory.

     

    It can be a tough learning curve, but I highly recommend this as a bible: Definitive Guide to DAX, The (2nd ed.) by Russo, Marco (ebook) (ebooks.com)

     

    Cheers,

    Adam