Hi,
I’ve more problems/understandings on my migration from LINUX dnsmasq to IPfire DHCP. My setings on IPfire for DHCP-Server. I offer only fix IP Addesses over MAC-Address.
I don’t understand why DHCP not distributed DNS- and Route Information for Clients. Clients ua. Raspian P4 Bookwurm. Useing networking.services. with this settings:
/etc/network/interfaces.d.eth0
Include files from /etc/network/interfaces.d:
source-directory /etc/network/interfaces.d/ #The primary network interface
allow-hotplug eth0
iface eth0 inet dhcp
I don’t understand this (on Client)
DHCPACK of 10.10.10.6 from 10.10.10.253
bound to 10.10.10.6 – renewal in 286 seconds.
Are you sure that you have no device with a TI lan or wlan chip in your network?
And have you only two devices?
You have limited the DHCP to use the addresses 10.10.10.100 - 10.10.10.101
if they are used there are no leases left that the server could assign.
On the raspberry pi: this is setup for DHCP on interface (iface) named eth0 (probably the on-board ethernet port)
DHCPACK is an acknowledgment from 10.10.10.253. But I do not know what 10.10.10.253 is. What is the IP of the IPFire box? Do you have any other DHCP servers like a cable modem or other router?
Please post a screen shot of you Zone Configuration page on IPFire. That should help decode this: unbound: [2190:0]|warning: duplicate local-zone Klara100.lan.|
DHCPDISCOVER from **38:81:d7:**43:f9:4c via green0: network 10.10.10.0/24: no free l eases. means you’ve run out of leases. You only have 2 available.
10.10.10.253. is IPfire address on green site ; and should also be the default gateway for all clients and the primary DNS Server for my Domain klara100.lan
Another point is; I use 3 IP addresses for one phyicall devices.
10.10.10.20 is
A.klara100.lan - Dataserver
B.klara100.lan - Nextcloud
C.klara100.lan - Datacloud
inside an IP network they are identified by an unique ( inside network ) IP address
IP addresses can be assign manually or automatically by DHCP
multiple application inside a device can be identified by a port number, it the task of the device’s network stack to transfer IP packets to/from the interface to the application
to deal flexible with device addresses, IP addresses can be associated to names; these are known on a logical level only ( DNS is your friend for name resolution )
If I interpret your description right, you have an device ( 10.10.10.20 ) which runs 3 applications. These are differentiated by port numbers. You can define multiple names for device. But this generates a n:1 relation between names and IPs. This relation can’t be used in reverse direction ( IP —> name ), so the DNS server ‘dices’.
thanks for your teaching me. Before I moved die DHCP-Server on IPfire
I use long time dnsmasq on LINUX; I’ve had no problems, with # Gateway
dhcp-option=3,10.10.10.253, no problems with fix address over dnsmasq ; no problems with DNS-Servises; no problems at all.
It seem that (IPfire) DHCP-Server is much difference to dnsmasq, but its a TCP/IP RFC Standard
device are … Layer 2 Protokol.
I confirm it
inside an IP network they are identified by an unique ( inside network ) IP address
I confirm it → TCP/IP Layer 3
multiple application inside a device can be identified by a port number, it the task of the device’s network stack to transfer IP packets to/from the interface to the application
I’ve used 1 Devies 2 VHOST (Apache-Server; NextCloud) with different names same IP-Adress and same Port 443; I use it since long a time;
Why does it works?
So, what does it mean for me ; It ist impossilble do match the dnsmasq config to IPfire DHCP-Server - right?
Conclusion back to dnsmasq - or? Because a can’t find any solution on your mail.
Maybe there will be practical tips and shows me my config error.
@checkpoint , sorry if my explanations were a bit ‘educational’.
But from your description I know not much about your ‘old’ config with dnsmasq, your projected config with IPFire, your network design.
To help in migration we should know the dnsmasq config adminst.
eth0 and eth1 have the same Mac address, so I suppose you’re using a USB switch. Probably a Texas well this might give you mixed results when having same IP’s. Some electrical jumps or a malfunctioning switch.
I thought the same when I first looked at it and that would have caused a huge problem if the mac’s of two nics were the same as that would cause all sorts of problem with the networking code in IPFire.
However looking carefully the last digit is different between the two MAC’s.
eth0 ends in c6 and eth1 ends in c7 so they are distinct MAC’s and should work fine.
G7, thanks; good idea; My Laptop has an USB-Ethernet Adapter and I take a look on it, but no success. No Texas MAC.
I keep searching; Not clear for me is
DHCPDISCOVER from 38:81:d7:43:f9:4c via green0: network 10.10.10.0/24: no free l eases
Texas Device is seaching over Broadcast on DHCP-Server in my LAN; I have only one (IPfire) and I offer in my I-range 10.10.10.100 - 10.10.10.130; but why takes it no a ip-address?
From my personal experience with Ethernet adapters and switches, I had to write on them with a Pen the Mac nr., or by mistake, I would plug the red on green and the Green on Red even on firstsetup when choosing adapters. It only takes one time to get it wrong. Changing the adapters after a fresh installation with setup can also lead to internal misconfigurations on resolvers. Mac’s aren’t reliable, my Windows still hasn’t found I’m using a different USB adapter since I changed from USB 2 to 3(+typeC) (one was Tplink other an AXIS) as well my HP Bios configuration allows me to choose a Mac of my choice. Some Cell phones have the function of hiding and/or randomizing Mac’s numbers too, so there might be some driver issues too due to vendors trying to fight Mac spoofing. I disabled DynamicHCP and assigned by hand devices but I have a very tinny green LAN.
*Are you on a virtual setup or it’s hardware instalation?
All MAC’s that start with 38:81:d7 are for items that have been manufactured by Texas Instruments. Of course that could be a network interface used in some other piece of equipment.
Do you have a switch, as mentioned by @g70p , that is either unmanaged and broadcasting generally or a managed switch that is trying to get an IP from your DHCP server.
If it is a managed switch then have you defined a static IP in the switch or have you set it to get an IP via DHCP. If the latter and you have not defined a fixed lease for it then you need to have more numbers available in your dynamic range.
Maybe you have changed the range but in your first screenshot in this thread your range was 10.10.10.100 - 10.10.10.101
If you have changed the Dynamic Range to go to 10.10.10.130 did you also press the Save button after updfating it as that is required to update the dhcpd.conf file with the changed values.
