Forum Discussion
Access lakehouse using logic app via Managed Identity
- 1 year ago
This is currently not feasible using copy jobs from Lakehouse to Sharepoint using Logic Apps.
Hi Anonymous
Thanks for the clarification.
I am facing another issue, when I copy csv, txt files its working fine. But when I copy the xlsx file its getting corrupted(not able to open file online or locally after downloading). I tried with multiple files, tried to convert the body to base64, binary before passing it as file content nothing seems to work.
I tried copying the excel file from lakehouse to blob, still it is corrupting the file. And I couldn't find any connector to 'get file content' from lakehouse.
can you please help me figure out why this might be happening. And how can I fix this?
Hi Anonymous,
Currently, when you use an HTTP GET request to the Lakehouse endpoint to read a file like .xlsx, the response is not automatically treated as binary in Logic Apps. It gets interpreted as text, which corrupts binary formats like Excel, images, PDFs, etc. So even though the file appears to download, it’s unreadable because it was never preserved in raw binary during transfer.
In Logic Apps, you can use "base64(body('HTTP'))" to convert the output to base64 or if you're passing it to another action (like SharePoint or Blob), map the raw body directly into the File Content field without converting to string/text.
For SharePoint or Blob target make sure the “File Content” field accepts base64/binary directly (depending on the connector). And in Blob storage, use "Create Blob" or "Upload Blob" with Content as body('HTTP').
And if you're using the SharePoint connector in the "Create file" or "Update file" actions, set File Content as "base64ToBinary(body('HTTP'))", if the connector expects binary input.
If I misunderstand your needs or you still have problems on it, please feel free to let us know.
Best Regards,
Hammad.
- Anonymous1 year agoNot applicable
I tried all these things and it still didn't work. so we decided to use blob storage instead of Lakehouse. Thanks for the guidence.