How can I share internet from one computer to another using ethernet? Both use Linux. The one with WiFi access has an Artix installation with openrc and connman. I think this is just a connman configuration but I have tried solutions on internet and it hasn't worked. * First, I tried enable tethering, adding ethernet to the configuration file in /etc/connman/main.conf. This works the best. It creates a "network" and let the other computer connect to it, without global internet connection, but it connects and has an IP (a static ip). I can even ping the other computer. But not reach internet connection. * I also tried lnxrouter, I installed dnmasq for it to work and apparently it does something it shows a PID and then says "Setting up completed, ... should be working". But it doesn't, I can't even connect (I think it was made for debian based distros only).

I'm running artixlinux with connman and wpa supplicant. I've never had any issues with other wifi networks. I can successfully connect to the wifi, and connman says 'Online' -but this only lasts for a few seconds - sometimes there's no internet at all. Below is the contents of /var/lib/connman/eduroam.config (I've taken the connman configuration from this archwiki page and I got the certificate file from the eduroam CAT ): [service_eduroam] Type=wifi Name=eduroam EAP=peap CACertFile=/etc/ssl/certs/Nottingham.pem Phase2=MSCHAPV2 Identity=someone@nottingham.ac.uk Passphrase=pass Sidenote - the eduroam CAT also generates a wpa_supplicant config file which looks like this: network={ ssid="eduroam" key_mgmt=WPA-EAP pairwise=CCMP group=CCMP TKIP eap=PEAP ca_cert="/home/user/.config/cat_installer/ca.pem" identity="someone@nottingham.ac.uk" altsubject_match="DNS:radius.nottingham.ac.uk" phase2="auth=MSCHAPV2" password="pass" } UPDATE - just figured it's probably a DNS issue because running ping 1.1.1.1 works. My /etc/resolv.conf looks like this: # Generated by dhcpcd from wlan0.dhcp # /etc/resolv.conf.head can replace this line domain nottingham.ac.uk nameserver 128.243.46.11 nameserver 128.243.42.5 nameserver 128.243.46.12 nameserver 128.243.42.6 # /etc/resolv.conf.tail can replace this line

I have a PC running K3OS with two ethernet interfaces. One (eth0) is connected to a network without a DHCP server with a static IP address and the other (eth1) is connected to a network with a DHCP, which returns a default gateway. K3OS uses connman as its network manager. I configured two connman services, c and d:

/net/connman/service/ethernet_00606561750c_cable
  Type = ethernet
  Security = [  ]
  State = ready
  Favorite = True
  Immutable = False
  AutoConnect = True
  Name = Wired
  Ethernet = [ Method=auto, Interface=eth0, Address=xx:xx:xx:xx:xx:xx, MTU=1500 ]
  IPv4 = [ Method=dhcp, Address=10.4.144.15, Netmask=255.255.255.0, Gateway=10.4.144.254 ]
  IPv4.Configuration = [ Method=dhcp ]
  IPv6 = [  ]
  IPv6.Configuration = [ Method=auto, Privacy=disabled ]
  Nameservers = [ 10.1.1.60 ]
  Nameservers.Configuration = [  ]
  Timeservers = [ 10.1.1.60 ]
  Timeservers.Configuration = [  ]
  Domains = [ domain.com ]
  Domains.Configuration = [  ]
  Proxy = [  ]
  Proxy.Configuration = [  ]
  mDNS = False
  mDNS.Configuration = False
  Provider = [  ]

/net/connman/service/ethernet_00606561750d_cable
  Type = ethernet
  Security = [  ]
  State = ready
  Favorite = True
  Immutable = False
  AutoConnect = True
  Name = Wired
  Ethernet = [ Method=auto, Interface=eth1, Address=xx:xx:xx:xx:xx:xx, MTU=1500 ]
  IPv4 = [ Method=manual, Address=192.168.10.3, Netmask=255.255.255.0 ]
  IPv4.Configuration = [ Method=manual, Address=192.168.10.3, Netmask=255.255.255.0 ]
  IPv6 = [  ]
  IPv6.Configuration = [ Method=off ]
  Nameservers = [ 8.8.8.8 ]
  Nameservers.Configuration = [  ]
  Timeservers = [  ]
  Timeservers.Configuration = [  ]
  Domains = [  ]
  Domains.Configuration = [  ]
  Proxy = [  ]
  Proxy.Configuration = [  ]
  mDNS = False
  mDNS.Configuration = False
  Provider = [  ]

Now what happens, after I reboot the PC or restart the service, I get a default route over eth1 set by connman. But eth1 doesn't feature a Gateway address so all my requests to other networks will fail. My understanding is that connman has an internal ordering of services. The ordering happens by respecting the settings of PreferredTechnologies in /etc/connman/main.config. When two services of the same technologies are available, it will priotize the one, which was ready the first (Source: Documentation ). But what I don't understand is, why connman would add a default route for networks without a default gateway. Is there a way to tell connman that a certain service does not feature a default gateway? Thanks.

I want to build a custom image of Slax and I am particularly annoyed by all those link-local (IPv4-LL) addresses it assigns to all interfaces by default. How can I get rid of that? I can't find any clues in man connman and man connman.conf.

The below route is sticky, even if manually deleted, it gets added by connmand (Connection Manager Daemon) with another interface name. If the specific interface is made down using ip link cmd, I see the same problem.

root@soc:~$ route -n
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
0.0.0.0         0.0.0.0         0.0.0.0         U     0      0        0 eth0

Is there a way to configure connmand not do this? Additional Information: eth0 uses static IP address. On my system (SoC), I have 4 interfaces, out of which 3 use static IP addresses and the one uses DHCP. I have added the systemd/network file that defines the static IP for eth0 below. root@soc:/etc/systemd/network$ cat 10-eth0-soc.network [Match] Name=eth0 [Network] Address=10.10.11.2/24 root@soc:/etc/systemd/network$

I am working on a Debian 11-based live USB distribution that utilizes Connman for network management. However, as this distribution may run on various computers with different network interface card identifiers, I am looking for a way to assign a distinct static IP to each machine. To achieve this, I plan to store the static IP in a file on each USB. To ensure network connectivity with the defined ips in a multi-instance environment, I would like to implement a mechanism that attempts to use the static IP upon booting the system. However, if there is no successful ping to an external server (e.g., google.com) or any other way to know the ip has been correctly leased, I want the system to fallback to DHCP for obtaining an IP dynamically. I would greatly appreciate any suggestions or guidance on how to implement this dynamic IP fallback mechanism effectively. Thank you.

I just finished configuring a fresh install of Arch on a Macbook Pro mid-2012 (9,2).
It has a Broadcom wireless modem (chip 4331) that works for internet browsing with my current configuration. SSH cliennt will not connect to any host, timing out. I have tested the same connection with two OS X versions, including one on this device, a shell emulator on my iPhone and an Ubuntu linux machine and all work. I am using connman and iwd to connect the wireless. Any suggestions would be much appreciated.

The rules for setting up an IP address on your computer in Linux has changed and now there is this connman thing. I'm on a Beaglebone black:

Linux beaglebone 4.19.94-ti-rt-r72 #1buster SMP PREEMPT RT Tue Mar 8 23:22:50 UTC 2022 armv7l GNU/Linux

I see there is a file,

cat /var/lib/connman/default.config 
[service_eth0]
Type = ethernet
IPv4 = 192.168.1.132/24

1) Do I change this file to change my IP address? For example, if I want my IP address to be 192.168.42.42, do I just change it in that IPv4 line? 2) If I want to change this to be assigned by the router (dhcp?), what do I do? 3) When does the change take effect? Do I have to reboot and/or stop and restart some service?

I run Debian 11 with the LXQt desktop environment. I have noted that Wi-Fi works with some routers but not with others. I also have the same software setup on a different computer and it has the same problem with the same routers. My conclusion is that it must be related to connman which is the default network manager in LXQt. The two routers I'm having problem with is an Apple Airport Express and my iPhone (used as personal hotspot). In the Connman window it says that I'm connected but still I have no internet access (see pictures below). I have also tried the command line: $ connmanctl Error getting VPN connections: The name net.connman.vpn was not provided by any .sconnmanctl> agent on Agent registered connmanctl> scan wifi Scan completed for wifi connmanctl> services [snip] *Ac iPhone wifi_d03c1f255599_6950686f6e65_managed_psk [snip] connmanctl> connect wifi_d03c1f255599_6950686f6e65_managed_psk Connected wifi_d03c1f255599_6950686f6e65_managed_psk connmanctl> quit $ ping www.google.com ping: www.google.com: Temporary failure in name resolution How can I troubleshoot the issue? What should I check next?

On a new debian 11 installation I get no internet, even when connman tells me that I am connected.

For a bunch of historical and justifiable reasons, I have a situation where I need to programmatically configure a device with a fresh Debian installation, over a ssh connection. Think of it like a production line where the device arrives fresh and needs to be configured over the network via a script. One of the things I need to do as part of that configuration is remove connman. If I do the most obvious thing and call apt purge connman from my script, it gets 20% the way through and the ssh connection is dropped. When I power cycle the computer and log back in, all is well, connman is gone and the network is configured the way I want. The user experience though, is a little less than desirable. Here are some workarounds I thought of and why they haven't been satisfactory: 1. echo a warning from the configuration script that the purge process will halt, and that the user needs to wait a period of time and then power cycle the device. * Not only is that a poor user experience, it means this has to be the last thing the script does, and I can't check for error or report success. 1. Require the user to plug in a keyboard and monitor instead of using ssh. * Pretty neat fix, but might not always be practical. 1. Do something like apt purge connman && reboot. * Difficult to provide user feedback; ssh needs to be re-established (in a different way because part of the configuration changes the network and ssh settings); takes an unpredictable while. 1. Schedule the purge for sometime in the future. * Neither at or crontab have a @before_reboot equivalent and a @reboot or other arbitrary time might be too late, can't report success, and requires cleaning up after itself so it only fires once successfully. 1. Configure connman to do nothing, rather than deleting it. * Leaves a lingering doubt that it is indeed inert. So is there a way to remove connman without killing the active ethernet connection? Can I prop up the connection until the next reboot, or just schedule the purge to be done during the next shutdown?

I have a Debian 8.7 IoT Version on a Beaglebone Black device which is connected to the Internet using a GSM Modem. I installed a VPN client (zerotier ) which joins a virtual network. Something changes the default route and my Internet is no longer working. netstat -rn Kernel IP routing table Destination Gateway Genmask Flags MSS Window irtt Iface 0.0.0.0 0.0.0.0 0.0.0.0 U 0 0 0 zt0 0.0.0.0 0.0.0.0 0.0.0.0 U 0 0 0 ppp0 10.64.64.64 0.0.0.0 255.255.255.255 UH 0 0 0 ppp0 10.241.0.0 0.0.0.0 255.255.0.0 U 0 0 0 zt0 169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 zt0 192.168.6.0 0.0.0.0 255.255.255.252 U 0 0 0 usb1 192.168.7.0 0.0.0.0 255.255.255.252 U 0 0 0 usb0 How can I ensure the default communication is routed over ppp0 and only 10.241.0.0/16 over zt0. After adding the default GW, I get the following Table: Destination Gateway Genmask Flags MSS Window irtt Iface 0.0.0.0 0.0.0.0 0.0.0.0 U 0 0 0 zt0 0.0.0.0 100.97.168.1 0.0.0.0 UG 0 0 0 ppp0 0.0.0.0 0.0.0.0 0.0.0.0 U 0 0 0 ppp0 I found out what adds the entry: Oct 02 12:55:48 PTMH7 connmand: zt0 {add} route 0.0.0.0 gw 0.0.0.0 scope 253 How can I prevent ConnMan from doing so?

So I wanna connect to a wifi network, I click it and it doesn't show the window to input the password. It just shows this error: "GDBus.Error:net.connman.Error.PassphraseRequired: Passphrase required". Same thing while using the connmanctl. I am on Garuda Linux Wayfire.

I have an embedded system with an operating system built by yocto (Automotive Grade Linux: Halibut 8.0) that has 2 competing dhcp clients: connman and systemd-networkd. sa8155:~# systemctl --type=service connman.service loaded active running Connection service systemd-networkd.service loaded active running Network Service I want to disable connman, but I get this strange error when I try: sa8155:~# systemctl disable connman.service Failed to disable unit: File /etc/systemd/system/multi-user.target.wants/connman.service: Read-only file system I am root. So I'm not sure why it will not allow me to delete this file. sa8155:~# ls -l /etc/systemd/system/multi-user.target.wants/connman.service lrwxrwxrwx 1 root root 35 Nov 27 2019 /etc/systemd/system/multi-user.target.wants/connman.service -> /lib/systemd/system/connman.service sa8155:~# Does anyone know why? What would be the correct way to disable this service? There seems to be an unusual number of file systems: sa8155:~# df -hT Filesystem Type Size Used Avail Use% Mounted on /dev/root ext4 1.4G 504M 812M 39% / devtmpfs devtmpfs 291M 0 291M 0% /dev tmpfs tmpfs 443M 0 443M 0% /dev/shm tmpfs tmpfs 443M 18M 425M 4% /run tmpfs tmpfs 443M 0 443M 0% /sys/fs/cgroup tmpfs tmpfs 443M 18M 425M 4% /etc/machine-id /dev/sde4 vfat 180M 53M 128M 30% /firmware /dev/sde5 vfat 64M 512K 64M 1% /bluetooth /dev/sde9 ext4 59M 21M 38M 36% /dsp /dev/sda2 ext4 122M 12M 108M 10% /var /dev/sda2 ext4 122M 12M 108M 10% /etc/bluetooth /dev/sda2 ext4 122M 12M 108M 10% /persist /dev/sda2 ext4 122M 12M 108M 10% /etc/build.prop /dev/sda2 ext4 122M 12M 108M 10% /etc/usb /dev/sda2 ext4 122M 12M 108M 10% /etc/misc/wifi /dev/sda2 ext4 122M 12M 108M 10% /etc/smack/accesses.d /dev/sda2 ext4 122M 12M 108M 10% /etc/adb_devid tmpfs tmpfs 443M 416K 442M 1% /var/volatile tmpfs tmpfs 443M 416K 442M 1% /persist/volatile /dev/sda8 ext4 108G 6.4G 101G 6% /data /dev/sda8 ext4 108G 6.4G 101G 6% /home tmpfs tmpfs 755M 4.0K 755M 1% /run/user/0 /dev/sda8 ext4 108G 6.4G 101G 6% /home/root/adas/dev /dev/sda8 ext4 108G 6.4G 101G 6% /data/root/adas/dev sa8155:~#

I'm trying to configure my new [*"pocketbeagle"*](http://beagleboard.org/pocket) to share a network connection with my Ubuntu (20.04) laptop. PocketBeagle runs a version of Debian: 10.3/2020-04-06 ([REF](http://beagleboard.org/latest-images)) . *"pocketbeagle"* has neither WiFi nor Ethernet - only a single USB port. It supports ssh, and I can access the board this way through either or both of two IPs. Perhaps easier if I show it like this: ### ifconfig run on *"pocketbeagle"* reveals:

$ ifconfig
usb0: flags=4163  mtu 1500
        inet 192.168.7.2  netmask 255.255.255.0  broadcast 192.168.7.255
        inet6 fe80::6264:5ff:fefa:813a  prefixlen 64  scopeid 0x20
        ether 60:64:05:fa:81:3a  txqueuelen 1000  (Ethernet)
        ...
usb1: flags=4163  mtu 1500
        inet 192.168.6.2  netmask 255.255.255.0  broadcast 192.168.6.255
        inet6 fe80::6264:5ff:fefa:813e  prefixlen 64  scopeid 0x20
        ether 60:64:05:fa:81:3e  txqueuelen 1000  (Ethernet)
        ...

### ifconfig run on *"ubuntu"* reveals:

$ ifconfig 
enx606405fa813b: flags=4163  mtu 1500
        inet 192.168.7.1  netmask 255.255.255.0  broadcast 192.168.7.255
        inet6 fe80::2267:92a0:a191:297d  prefixlen 64  scopeid 0x20
        ether 60:64:05:fa:81:3b  txqueuelen 1000  (Ethernet)
        ...

enx606405fa813d: flags=4163  mtu 1500
        inet 192.168.6.1  netmask 255.255.255.0  broadcast 192.168.6.255
        inet6 fe80::3898:bc86:d0c2:cb53  prefixlen 64  scopeid 0x20
        ether 60:64:05:fa:81:3d  txqueuelen 1000  (Ethernet)
        ...

### Through a bit of research, I formulated a procedure to effect sharing as shown below: 1. Using the *"ubuntu"* GUI, I changed the *"Ethernet Connection"* w/ MAC 60:64:05:FA:81:3D from its default *"automatic DHCP"* to a *"shared connection"*. 2. This changed its IP address from 192.168.6.1 assigned by *"pocketbeagle"* to 10.42.0.1. After making this change, here's what the *"ubuntu"* reports on its CLI:

$ ifconfig
enx606405fa813d: flags=4163  mtu 1500
        inet 10.42.0.1  netmask 255.255.255.0  broadcast 10.42.0.255
        inet6 fe80::3898:bc86:d0c2:cb53  prefixlen 64  scopeid 0x20
        ether 60:64:05:fa:81:3d  txqueuelen 1000  (Ethernet)
        ...

3. At this point, the interfaces on *"pocketbeagle"* are unchanged, but obviously incorrect for connection sharing. I executed the following from the CLI on *"pocketbeagle"*

$ sudo connmanctl config usb1 --ipv4 manual 10.42.0.2 255.255.255.0 10.42.0.1 --nameservers 10.42.0.1
[sudo] password for debian: 
Error usb1: Method "SetProperty" with signature "sv" on interface "net.connman.Service" doesn't exist

Attempting a query of services w/ connmanctl:

$ sudo connmanctl 
Error getting VPN connections: The name net.connman.vpn was not provided by any .service files
connmancconnmanctl> services

connmanctl>

I'm not at all familiar with connmanctl, but my research suggested it was the only way to properly set up DNS service. Perhaps I screwed up? Perhaps there is something else I've missed? I then decided to try the configuration the "old way": 4. The following sequence of commands was used to re-configure the corresponding *"pocketbeagle"* interface in the "old way":

$ sudo ifconfig usb1 10.42.0.2 netmask 255.255.255.0 up
$ sudo route add default gw 10.42.0.1 usb1
$ sudo nano /etc/resolv.conf
#
# add line: nameserver 10.42.0.1 
# save & exit editor
#
# yes - this is an awful kludge, but it did provide DNS

**And this worked!** - the *"pocketbeagle"* successfully connected to the Internet, DNS worked also.

debian@beaglebone:~$ ping unix.stackexchange.com
PING unix.stackexchange.com (151.101.1.69) 56(84) bytes of data.
64 bytes from 151.101.1.69 (151.101.1.69): icmp_seq=1 ttl=56 time=9.01 ms
64 bytes from 151.101.1.69 (151.101.1.69): icmp_seq=2 ttl=56 time=11.8 ms
...

### QUESTION: Obviously the issue is with my use of connmanctl, and not with *"pocketbeagle"* itself. What did I do incorrectly - why will connmanctl not work as intended?

We are looking for a way to get ConnMan to hash passwords (hashed on disk, i.e. not in plain text). We are wondering if it is possible to enable such feature by compile-time definition or runtime flags Could you please advise us on this?

I'm using my Laptop on several networks. When I'm at home or at the office I occasionally want to connect with my phone to a server running on my laptop (say MPD, ssh, transmission). The problem I'm facing at the office is that my IP address changes often. Hence I need to change the IP the apps on my phone use to connect to the servers and it's pretty tedious. I don't have permission to configure the router at the office so it'll give a static IP specifically to my machine, as I've done at home. What's more frustrating, is that the office's router doesn't seem to support local network DNS - my machine's name is NUX and typing e.g ssh NUX on my phone while at the office fails. I've often heard about a static IP address configured on the machine and not on the router but I never figured out how to do that. If I'd like to configure that wouldn't the router expect me to want a static IP? It doesn't make sense that a router configured for DHCP will go along with a machine demanding a static IP. Besides setting a static IP with such a router, I'm considering writing a dirty hack to workaround the issue: Run a script on startup that will parse the output of nmap -sP 192.168.1.0/24 and send some kind of message to the phone indicating it's own IP address. I'm using [connman](https://01.org/connman) on [NixOS](https://nixos.org/) but I'm willing to consider an alternative network manager for the sake of making this work.