Maintaining EC2 Servers with Chef Server

I have a hosted chef account working with AWS EC2 instances.

The instances can connect to hosted chef by way of chef-client and run the inital run_list that is presetly a role called servers. Everything runs and installs correctly on initial provision.

I should also mention I have autoscaling on ec2 instances that increses the amount of instances provisioned on an 'as needed' basis and this works fine.

I found one talk that sort of makes sense but seems a bit more involved. http://ift.tt/1uW7Doq

My questions are around how to maintain the servers after initial provisioning. First, if I have an update to my role to include more cookbooks or configuration, and how would I push this out to say X amount servers in a simple way.

Second, how would chef server know if a node has failed and notify the admin?

Any guidance?