Airflow mysql to gcp Dag error

Question

I\'m recently started working with Airflow. I\'m working on DAG that:

1. Queries the MySQL database
2. Extract the query and stores it in a cloud storage bucket

Answer 1

According to your traceback, your code is breaking at this point. As you can see, it process the code:

json.dump(row_dict, tmp_file_handle)

tmp_file_handle is a NamedTemporaryFile initialized with default input args, that is, it simulates a file opened with w+b mode (and therefore only accepts bytes-like data as input).

The problem is that in Python 2 all strings are bytes whereas in Python 3 strings are texts (encoded by default as utf-8).

If you open a Python 2 and run this code:

In [1]: from tempfile import NamedTemporaryFile
In [2]: tmp_f = NamedTemporaryFile(delete=True)
In [3]: import json
In [4]: json.dump({'1': 1}, tmp_f)

It works fine.

But if you open a Python 3 and run the same code:

In [54]: from tempfile import NamedTemporaryFile
In [55]: tmp_f = NamedTemporaryFile(delete=True)
In [56]: import json
In [57]: json.dump({'1': 1}, tmp_f)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
 in ()
----> 1 json.dump({'1': 1}, tmp_f)

/usr/local/lib/python3.6/json/__init__.py in dump(obj, fp, skipkeys, ensure_ascii, check_circular, allow_nan, cls, indent, separators, default, sort_keys, **kw)
    178     # a debuggability cost
    179     for chunk in iterable:
--> 180         fp.write(chunk)
    181 
    182 

/usr/local/lib/python3.6/tempfile.py in func_wrapper(*args, **kwargs)
    481             @_functools.wraps(func)
    482             def func_wrapper(*args, **kwargs):
--> 483                 return func(*args, **kwargs)
    484             # Avoid closing the file as long as the wrapper is alive,
    485             # see issue #18879.

TypeError: a bytes-like object is required, not 'str'

We get the same error as yours.

This means that Airflow is still not fully supported for Python 3 (as you can see in the test coverage, the module airflow/contrib/operators/mysql_to_gcs.py is not yet tested either in python 2 or 3). One way to confirm this would be to run your code using python 2 and see if it works.

I'd recommend creating an issue on their JIRA requesting portability for both versions of Python.