Missing fail security

[ftrueck]

The docker container crashes if something goes wrong.
In terms of fail security it would be better if the tasks fail gracefully with a chance of recovery so that you do not need to start your docker every time an exception occours.

Examples:

• if you send wrong commands via mqtt the script exits and the container stops
• if a device (ac unit) is not reachable via network the script crashes
• if a DNS resolve fails the script crashes
• sometimes the script crashes when a command gets send double in too short amount of time

These kind of errors should not crash the entire script as there may be other units that are still functional and can be operated.

The solution would be a bunch of try-catch directives to protect the main thread of the application to remain operational.