create_str = "CREATE PROC sp_General_GetAllData (@table_name AS NVARCHAR(MAX),
              @year AS INT, @month AS INT, @pd AS INT)
              AS
              BEGIN
                  SET NOCOUNT ON;
                  DECLARE @sql NVARCHAR(MAX) # for later uses
                  IF @table_name LIKE '%[^a-zA-Z0-9_]%'
                  BEGIN
                  # the 'log' table allows NULLs and the fields are used in other cases
                      INSERT INTO [myDB].[mySchema].[log]
                      VALUES (SUSER_SNAME(), NULL, NULL, NULL, NULL, NULL,
                              'INVALID TABLE NAME: ' + @table_name, GETDATE())
                      RAISERROR ('Bad table name! Contact your friendly DBA for details', 0, 0)
                      RETURN
                  END
              # do some things if the @table_name is ok...
              END"
cursor = sql_conn.execute(create_str)
cursor.commit()

# calling the SP from python - doesn't write to the log table which get stucked
query = "{CALL sp_General_GetAllData (?, ?, ?, ?)}"
data = pd.read_sql(query, sql_conn, params=['just*testing', 2019, 7, 2])

# calling the SP from SSMS - works fine
EXEC sp_General_GetAllData 'just*testing', 2019, 7, 2