The DHCP server from Jack prabha's blog

The DHCP server must be configured with an appropriate scope of IP addresses and TCP/IP parameters for clients within each subnet. For example, the administrator must configured the DHCP server with an IP address range and subnet mask. The administrator must also set the lease duration time dependant on the organisations requirements. For example, if the organisation only has a limited number of IP addresses available and a large number of devices or clients that join and leave the network regularly, then the administrator may wish to reduce the time of the lease so that DHCP returns unused addresses for redistribution. 

Read more: One Stop MOS

Finally, the administrator should configure the DHCP server to exclude static IP addresses present on the network from scope. For example, you may choose to exclude a certain IP address range from the scope of DHCP and utilise this address space for statically assigned servers on the network, if appropriate.

The interactions between client and server can be broadly compared to that of a TCP/IP connection. The client utilises a DHCPDiscover command to search for the DHCP server on the network. This is achieved by broadcasting the message to a specific subnet where the servers reside. The DHCP Server should pick up this command and issue a DHCPOffer command back to the client, which offers an IP address as configured by the administrator. The client then responds with a DHCPRequest command that requests the IP address offered and finally the server issues a DHCPPAck message to finalise the terms of the lease. This can be easily compared with shopping or an auction, as the seller you put out your product to a specific market, a potential buyer see’s the advert and comes back with an “offer” and you as the seller accept or decline and then the sale is finalised by the buyer presenting cash.