One Nic can be configured with multiple IP addresses. Only one IP address in the same subnet is the primary IP address, and all IP addresses in the same subnet are the secondary IP addresses. The primary and secondary IP addresses are allocated for convenient management, the primary and secondary addresses exist for two purposes, one of which is to build different subnets on the shared media.ArticleThis situation is described in the differences between ip addr add and ifconfig to view the IP address structure of the Linux Nic. The network of this shared media is a very messy network, the standard requires that data must pass through the router across subnets, but it is a waste of time to finish this on the shared media, so there is a "redirection" type ICMP packet, this troublesome operation is limited to ARP. After ARP, the router delivers a redirected ICMP packet to the source host. We recommend that the source host directly communicate with the target host, this is a typical use of existing standards to cater to new standards, IP in Ethernet communication can not hide ARP, but ARP and ICMP work together to make the routing change, the lower layer can also affect the upper layer, instead of simply carrying the upper layer, the other place that uses the Primary and Secondary addresses is to put IP addresses and multiple applicationsProgramBinding. Each application needs a different IP address.