Stefano,
I am not sure whether I communicated correctly:
I have a three-tier set-up here. My Anvil App front-end communicates with my server code through uplink. Thus all my server code executes on my local machine that is in turn interacting with the Azure DB as backend.
I only maintain minimal login/user data on Anvil Server.
I cannot/don’t want to develop in the online editor due to lack of debugging and other tools (mainly SQL-console).
I use PyCharm instead. Already client development can be tedious in the online editor.
There is, furthermore, the not insignificant problem that if you want to access Azure from another cloud location (ie in server code in Anvil) you run into firewall/authorization problems on the Azure side.
My (maybe stupid) plan is to multiply my servers that interact with Azure.
In PyCharm, I use the pyodbc driver to interact (it is Microsofts version). That seems to work pretty well.
I appreciate any suggestion that improves the scalability of my app!
Franz