I have a customized Embedded Linux image made using Yocto. The single board computer on which the image is loaded is default configured with a static IP address of where I can set a static IP address on my host machine and connect to the board directly to the single board computer (target) via SSH and the static IP address (

I have found a file for the systemd network configuration as following:

# Path /lib/systemd/network/10-eth0.network


According to the systemd-networkd docs I can understand that the DHCP=ipv4 is to accept the IPv4 Addresses when connected to a router. But I do not understand the Address after it. Is it a fallback to static IP address?

I do not have any other way to reach the board except SSH for the board. If I connect it to an infrastructure LAN where I have no way to reach the router to determine the IP address via DHCP and would like the board to fall back to the static IP address.

Is this achieved via the script above?

1 Answer 1


I just discovered that dhcpcd has this -S option for static assignment... It works if I use it twice, once for ip_address= and once for routers=.

I also can connect through ip link, ip address and ip route.

With systemd I am very careful. And your example shows why: it is a complex layer of configuration and rules.

The Address= means static instead of dynamic, as I see, just like the -S option above. But the filename "10-eth0..." is also important. Has to do with udev according to systemd, just like the "match" rules. Together with other files this could well lead to some "fallback" config...

btw your link is to the arch wiki. systemd has good docs itself, but it is just too much for me!

The systemd docs explain somewhere that the simple "network is up" variable or status is too simple.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.