The following does not work:

1. Reading from a database into a pyspark df
2. Performing some modifications on df
3. Overwriting that same database with df

In the end, the databse whould contain the updated version of df, but it is always blank, as the attached code snippet illustrates.