After a series of errors connecting to message router, the distribution client just stops attempting to get and process events.  It just stalls.  My application that uses the distribution client keeps running the distribution client because no error was thrown.  The application becomes a zombie.

The attached log stopped at "18-04-18 18:04:12" but my app is still up and running

$ date -u
Thu Apr 19 20:35:44 UTC 2018

What would be better is if the distribution client either:

1.  Throws an error to the application to notify the application and let the application handle it either re-creating the distribution client or just killing the app

2.  Require the application a callback to handle errors.  This is less attractive than the first because there's no way to do more complex error handling like retries.

3.  Kill the app.