Python - good practice to catch errors

Question

I have these files in my project:

- main.py
- module1.py
- module2.py

main is the main file that will be directly execute in console.

module1 will be imported into main and throws Module1Exception.

module2 will be imported into module1, throws Module2Exception and uses lib that throws NormalException and CriticalException exceptions.

On all exceptions application should exit.

Which option is a most effective way to catch those errors?
a) All exceptions will be catch in main, print message and exit
b) Lets allow modules to catch exceptions (module2 catch lib exceptions, module1 catch only Module2Exception and main catch Module1Exception), print message and exit.
c) Like b, except that modules will not exit, instead returns False and exit will be called in main

Edit:

This will be server side application, running as daemon. I expect that in future I will be using multiple servers with global log monitoring, like flume or scribe. Exceptions must be caught. These files is only example. In fact, I'm writing a large application which is act as a kind of server. In this moment i write module for load and parse configuration file. In this case script should exit if any exceptions will be raised. And only in this case. After loading the configuration files and pass the tests, the script will be left unattended.

Answer 1

main should have the responsibility to terminate, not the modules.

What if you wanted to import those modules into another project and you didn't want to terminate the application on an exception?