I'm hoping I can get some pointers to get some niggles in my workflow resolved.
Getting tableau server talking to the SQL server of interest is pending so for now I'm trying the following in python:
Query server>write results to .hyper file (using Tableau Extract API)>Publish .hyper to server(using Tableau Server API)
With the intent to then connect to the .hyper file with workbook to generate a dashboard, and then ongoing to be able to push updates to the .hyper file that will appear in the dashboard with a scheduled python script.
The hyper file is created fine, if i connect to the local file with tableau desktop I can read in the data from the table. I can then publish the file from desktop to server. I can then use the in browser editor to create a dashboard from the datasource published from desktop.
However if i try and connect to the datasource that's been published through the API in python the table doesnt appear to be accessible. Opening the datasource with the in browser editor doesnt show the fields and shows nothing in terms of options to select the table within the hyper file. Similarly attempting to connect to the datasource on tableau server through tableau desktop fails. It loops on a "recconect to [datasourcename]" message without loading any data or providing an option to select the table from within the file like you get when connecting to the local file.
What might I be doing wrong when publishing the hyper file through the server API thats rendering the contained data not visible when connecting given that the file is fine when viewing in desktop?
Think I got it, the tables have to be called "Extract", other table names cause the above issue. Not sure if i missed this in the docs or if it wasnt mentioned. Discovered this as a limitation mentioned in someone elses df to hyper library.