We are maintaining more and more end users with 3-node clusters in load sharing mode. One R80.10 end user has activated the DHCP Server on his three clusternodes and therefore receives dhcp duplicates on his systems caused by all three clusternodes actively responding to dhcp requests.
ClusterXL Load Sharing mode limitations and important notes does not mention any existing limitation for DHCP Server.
The end user could send DHCP requests to the physical IP address of one of his three clusternodes only, but this would mean a loss of redundancy for his DHCP functionality.
WHat would you suggest?