For embedded systems, I would recommand connman over networkd exactly
for that sort of reasons.
networkd has been designed mainly for the datacenter use-cases with lots
of interfaces appearing and disapearing dynamically (dynamic containers)
for the embedded case where we have only a few network interfaces but
the network itself changes a lot, connman might be a better choice.
Post by Benjamin Wozniak
we are developing an embedded device that can only be accessed via network.
If DHCP is used for network configuration we need a plan B if for some reason
no lease can be obtained. Our idea is to have a static IP as fallback.
In systemd I didn't find any hooks to react to a lease fail. The files in
/var/run/systemd/netif apparently should not be used ("# This is private data.
Do not parse.").
Is the idea of a fallback IP address just not a good one, is such functionality
just not implemented in networkd or am I missing something here?
Thanks in advance,
systemd-devel mailing list
20 rue des Jardins
Responsable de l'expertise Smile-ECS
email ***@smile.fr <mailto:***@smile.fr>
Twitter <https://twitter.com/GroupeSmile> Facebook
DÃ©couvrez lâunivers Smile, rendez-vous sur smile.eu
eco Pour la planÃšte, n'imprimez ce mail que si c'est nÃ©cessaire