Client computers get their DHCP Server Setting

Technology Microsoft
1

Where do client computers (Windows 10) get their DHCP Server setting from? when you do an ipconfig /all. I know it’s set on the DCs but not sure where.

2

DHCP is Discovered as part of the protocol. Specifically, when a device connects to a network it sends a broadcast DHCP Discover message out, and whatever responds is what it goes with.

3

Right. i have two DCs and both are DHCP servers. i was wondering how one of them gets picked over the other during discovery.

4

Whichever is the first to respond if they are both configured with the DHCP service.

1 Like
5

If you are running two DHCP servers on your network, you’re likely going to want to configure correct High Availability so that they each work together and sync the leases, otherwise the behavior can be a bit erratic.

https://blogs.technet.microsoft.com/teamdhcp/2012/06/28/ensuring-high-availability-of-dhcp-using-windows-server-2012-dhcp-failover/

6

thank you. yup. i’ve got HA correctly configured. I’ll take a look-see at this article to verify

7

In an HA setup, there’s some coordination that goes on but the basics of it are they split the pool (say, for two servers, they’d split it 40% Server A, 40% Server B, 20% Reserve) and they’d exchange heartbeats back and forth to make sure they’re both healthy and happy. They also inform each other of leases as they are issued. Then, when everything is good, they’ll both see the DHCPDISCOVER request, take the MAC Address and see if it’s even or odd (the actual math is slightly more complex, but for the example that’s fine) and the server responsible for those MACs will answer back and complete the process.

If one of the servers goes offline gracefully (via a reboot or whatever) it will tell the other server “I’m going offline, please handle all DHCP for now” and the remaining server will behave as a normal DHCP server. If it goes down unexpectedly, you’ll see some delays for some clients connecting while the heartbeat timeout occurs (usually not that noticeable) and then the remaining server will issue an IP out of the Reserve pool. After an extended period of no heartbeat, the remaining server will fail back to standalone operation.

8

Great explanation. thank you so much