Forum Discussion
PowerBI with Azure Data Lake Store
I found out that ir you are looking to use an Azure Data Lake as your data store, you cannot directly connect to it online. If you want to use an Azure Data Lake as a data source, you will need to use an ODBC data source in Power BI Desktop.
I managed to get data from a Data Late Store into PowerBI desktop, but not the way I would like it:
- PowerBI Desktop, Version: 2.32.4307.641 64-bit (February 2016)
- Get Data / Azure / Microsoft Azure Data Lake Store (Beta)
- Provide the adl:// or swebhdfs:// path
- It will show you a preview with columns like "content, name, extension, .."
- Click on "Edit"
- Click on the link "Binary" in the first column ("Content")
- Now you can see the the data as it appears in the file itself.
That's a very interesting solution. Thanks for sharing. Here's a question, are you able to refresh the data extracted from the datastore in the personal gateway?
I published the report to PowerBI.com and I could refresh the data there. Don't need the gateway, there's no local to cloud data movement involved here. (Direct connection). I just needed to re-enter my credentials for the DataLake reference.
I am getting the followin error, while trying to re-enter the credentials
Failed to update data source credentials: The credentials you provided for the data source are invalid. Please check the credentials for all the data sources.Hide details
Activity ID: f3753c07-f1f2-4d8e-b55e-af0a9b0ec4e8
Request ID: 323a3391-1a49-6c15-3ed8-c0792e4c2610
Status code: 400
Time: Wed Nov 16 2016 13:26:42 GMT+0100 (W. Europe Standard Time)
Version: 13.0.1700.581
Cluster URI: https://wabi-north-europe-redirect.analysis.windows.net
Details: The credentials provided for the DataLake source are invalid. (Source at https://xxx.azuredatalakestore.net/yyy.)
What happens when we have files stored in multiple sub folders by timestamp?
When I browse binary data and check the scripts inside Power BI advance editor I get follow:
let
Source = DataLake.Contents("https://xxxxxxxx.azuredatalakestore.net"),
Content = Source[Content],
Content1 = Content{0},
#"2017" = Content1{[Name="2017"]}[Content],
#"04" = #"2017"{[Name="04"]}[Content],
#"20" = #"04"{[Name="20"]}[Content],
#"14_0_f44e6b5c95fa49dfa3b78c89878777d1 json" = #"20"{[Name="14_0_f44e6b5c95fa49dfa3b78c89878777d1.json"]}[Content],
#"Imported CSV" = Csv.Document(#"14_0_f44e6b5c95fa49dfa3b78c89878777d1 json",[Delimiter=",", Columns=17, Encoding=1252, QuoteStyle=QuoteStyle.None]),which tells I went all the way to 2017-04-20 and picked that day file.
Can I setup this to load all files as they come?
Shabbir
how can you we load all the files for all the files.. I am not sure clicking Binary would work if I want to work with all of the data?
Would it
