DHCP High Availability can be configured in multiple ways to handle failure scenarios and load balancing. In all high availability scenarios, ensure that the two or more DHCP servers are on separate hosts. We can accomplish this by storing the DHCP server on local disk or with DRS anti-affinity rules. DHCP servers are often hosted on domain controllers as the resource requirements are quite low. DHCP needs to be separated from the domain controller role in the largest environments only.
Active/Backup: In this mode, we configure a primary DHCP server to handle all the DHCP traffic on the network. The backup node will receive DHCP state information but will not handle any DHCP traffic while the primary is active. Once the primary DHCP server fails, the backup takes over. This can be accomplished with server clustering solutions such as Windows Server Failover Cluster (WSFC) or by using Windows Server 2012's built in DHCP clustering.