default dhcp - If checked, this scope will be used for clients when their network can't be determined.

[Simon-CR]

Hi,

so I have multiple scope configured handing out IPs without issues (all via relay).

I enabled the "If checked, this scope will be used for clients when their network can't be determined." on one of my scope. All of a sudden dhcp client that had IP addresses from their respective scope started getting IPs from the "default" enabled scope instead.. turning it off returned things to normal

[BeryJu]

hmm, yeah the intended functionality of this is that "as a last resort", if scope can't be detected via network/relay then the scope with this flag enabled should be used, it shouldn't affect correct matches to other scopes

[Simon-CR]

what are the criteria for a scope to be detected via network/relay? all my configured subnet / scope are using the same relay server that throws the request to gravity.

[BeryJu]

https://github.com/BeryJu/gravity/blob/main/pkg/roles/dhcp/scopes.go#L107-L128 is the logic to find a scope for a DHCP request

[tompaah]

I'm having an issue similar to this. Three VLANS with ip-helper configured (dhcp forwarder by Cisco terms), thus three DHCP scopes (named: container, container-dev, ilo). None of them are marked as "Default".

I made a "Reservation" for one machine in the "ilo" scope. When fetching a DHCP adress, it got an address from the "ilo" scope. So far so good.

But next time, when renewing the address (TTL is set to 3600, it renewed after approx. 1/2 hour so I guess renewal normally happens after 1/2 the time has elapsed) it got an adress from the "container" scope and thus became unresponsive on the network (wrong IP adress for the subnet).

Can I provide anything in form of logs or other stuff?