Wed Dec 04, 2024 2:55 pm
Client ID is something DHCP clients supply to server (and not the other way around). When DHCP server decides on which lease to offer, it first checks "Client ID" ... and only if that value was not provided by client, it falls back to using client MAC address (as identifier). Indeed most DHCP clients construct "Client ID" using own MAC address, but it doesn't have to be. BTW, "active host name" is also provided by DHCP client in ROS case (ROS DHCP server by default doesn't send Option 12 (Hostname) to clients).
Regarding "unidentified network": it's windows internal stuff and most probably it's based on several information pieces (including but not limited to IP and MAC addresses of default gateway). Based on this information, windows asks user which kind of network it is (home / work / internet) and adjusts windows firewall settings accordingly. Next time windows encounters same set of network information pieces, it declares network as identified and applies same firewall settings as they were first set.
Now, if many windows clients complain about unidentified network and you know that same windows machine was connected to same network previously and you know that none of information about LAN (used by windows to identify network) is changed, then you may have some problem in your network ... e.g. rogue DHCP server.
Just a comment on screenshot, posted by @OP: I've no idea how DHCP server was able to send out a dynamic lease without recording client MAC address and ClientID (row, pointed at by red arrow).