Forum Discussion
600 Second Timeout Setting?
etuckeriv - Are you really sure that you need to download Analysis Service cube data into Power BI model? You could be better off using SSMS or DAX Studio to integrate the data, or importing the data from Source. If from Power BI dataset, extract too much data could exceed CPU and memory limits leading to poor performance.
The following will increase the timeout from 600 seconds to 900 seconds. Note #duration( day, hour, minute, seconds)
Source = AnalysisServices.Database( #"Server", #"Database", [Query="MDX or DAX String", ConnectionTimeout=#duration( 0, 0, 15, 0 ) ] )
The top recommended post to the right has the solution:
Function Odbc.DataSource error with Timeout settings
As mentioned in the documentation, you can use "ConnectionTimeout":
Source = Odbc.DataSource("dsn=Production", [HierarchicalNavigation=true, ConnectionTimeout=#duration(0, 0, 15, 0)])
Thank you so much, I didn't realize those were over there! I will definitely pay more attention to that area in the future.
Also, thank you for the suggestion! I am having trouble figuring out where to put that line. I looked through the post you linked (as well as several others that seemed related from the side bar), as well as the docmuentation you linked, but I don't fully understand what I'm seeing in the other solutions and it looks different from mine. I also tried adding the line or pieces of it in various places within the query code, but everywhere I tried I got an error that it wasn't valid. I realize you can't see what I'm seeing, I'm nervous about sharing the query code or a screenshot because I'm not sure what information might be sensitive. Would you be willing to suggest where I should put the line of code? Or what info you might need to be able to tell me what to do?
Usually, pasting the M code from the Advanced Editor works well provided you change or redact any names or other sensitive bits that might be within filters or column names since the query just says how to get and transform the data rather than what the data actually is.
