Hi. In this case, you point traffic to the primary IP address. If the primary (220.127.116.11) fails and the secondary (18.104.22.168) takes over, the secondary will adopt the IP (22.214.171.124) of the primary, and continue forwarding like nothing happened (assuming you have a link state interface in your failover config to sync the state table). I believe the secondary in this case sends a gratuitous ARP or maps the MAC address of the primary to the secondary in the process to facilitate connected devices from getting into a WTF situation. Note that this is basically how VRRP works, but I dont believe the ASA is running VRRP, but some proprietary mechanism.
I hope that helps (I hope it is correct, also)