Forum Discussion

stacey_k's avatar
stacey_k
Frequent Visitor
7 years ago

Poor performance on Power BI Service

I am building a fairly large report using imports and scheduled refreshes with some pages containing up to 18 cards and 5 graphs. There are 7 pages and 24 queries in total, but all from the same source or manual data entries

 

This seems to have significantly affected the report's performance despite the pbix file being only 5MB in size.

 

Trying to navigate the report on the Power BI service is very difficult as a full page sometimes takes up to a minute to load all the visuals. Switching between tabs has a slow response time and also seems to re-load all the visuals on the page at every switch.

 

I've tried to align the report to Power BI Performance Best Practices as far as possible (with respect to requirements) but I'm not seeing much improvement.

9 Replies

  • Hello, stacey_k , are you using Import Data or Direct? 

     

    It seems you are using Direct, because, when you importing data and publishin into Service, you will have a high and fast perfomance to slicer and load visuals.

     

     

    • stacey_k's avatar
      stacey_k
      Frequent Visitor

      Definitely using Import as per below:

  • Anonymous's avatar
    Anonymous
    Not applicable

    I too am noticing an increasing number of timeouts and failures in recent weeks on scheduled refreshing of reports that pull data from the on-prem Jira instance via an enterprise gateway.

     

    Typical errors are:

    Hide
    Scheduled9/25/2018, 2:01:27 PM9/25/2018, 2:38:47 PMFailedTimeout occurred when sending or receiving a request.
    Cluster URI:WABI-US-NORTH-CENTRAL-redirect.analysis.windows.net
    Activity ID:46dc7a06-0771-4bc8-8829-b211310a761f
    Request ID:d015bffa-e32c-4f7c-9810-1a006e2e63d8
    Time:2018-09-25 19:38:47Z
     

     

    and 

     

    Details Type Start End Status Message

    Hide
    On demand9/26/2018, 12:05:36 PM9/26/2018, 12:25:57 PMFailedCannot connect to the mashup data source. See error details for more information.
    Underlying error code:-2147467259 Table: Story Burndown.
    Underlying error message:The operation has timed out
    DM_ErrorDetailNameCode_UnderlyingHResult:-2147467259
    Microsoft.Data.Mashup.ValueError.Reason:DataSource.Error
    Cluster URI:WABI-US-NORTH-CENTRAL-redirect.analysis.windows.net
    Activity ID:f4517f00-cfb7-4b26-96d4-7399e040f795
    Request ID:dff9992d-203e-cb7e-48a5-c12d3e50b99e
    Time:2018-09-26 17:25:57

     

  • v-jiascu-msft's avatar
    v-jiascu-msft
    Microsoft Employee

    Hi stacey_k,

     

    Is the performance fair enough in the Desktop? Since you just consume the report when it slows down, I guess there are some measures that have heavy calculations. Can you share the file?

     

     

    Best Regards,
    Dale

    • stacey_k's avatar
      stacey_k
      Frequent Visitor

      Performance is fine on Desktop - can't share the file unfortunately due to IP concerns. Any suggestions on how I can tackle this?

      • Anonymous's avatar
        Anonymous
        Not applicable

        I have the same observation - desktop performance is consistent.  Report refresh through the service/gateway has degraded over time and is experiencing an increased failure rate due to timeout or mashup data connection in recent weeks. There have been no notable changes to the report in several months.

  • v-jiascu-msft's avatar
    v-jiascu-msft
    Microsoft Employee

    Hi stacey_k,

     

    It could be a temporary issue. How about now? If it continues, please provide the information about the location of Power BI. Please refer to the snapshot below.

    Poor_performance_on_Power_BI_Service

     

    Best Regards,
    Dale

      • v-jiascu-msft's avatar
        v-jiascu-msft
        Microsoft Employee

        Hi stacey_k,

         

        Not heard any issues in that cluster. I would suggest you create a support ticket here if the issue persists.

        create ATicket

        Best Regards,
        Dale