No ideas why the NS setup kills the network? The problem looks a little bit like those after installing NS on top of Centos in a LXC container.
Unfortunately my board runs headless. So I can only connect through ssh or I have to mount the emmc chip on a different PC in order to read the log or to manipulate any files.
Oke this complicates things…
Did you have a fixed or a DHCP IP before you ran nethserver-install?
Maybe you are bitten by the same thing which I though was specific to the RPI-4.
If it was DHCP you may salvage your install by editing either the eth0-icfg script or manualy creating a simlink to enable networkmanager. For the latter you probably need to chroot the ROOTFS on your emmc chip.
I would try to go with editing ../etc/sysconfig/network-scripts/ifcfg-eth0 (assuming the network interface defaulted to eth0) to kickstart the network and provide something of the like off :
in which w.x.y.z needs to be a valid ipadres in your network.
NOTE: this hack will be over written if you change the network configuration in nethserver.
So if you succeed provide a valid (FIXED-IP) setup in ns asap
BTW: I can highly recommend to get a hardkernel serial-to-usb converter for headless tinkering
Nope. Your hint looked promising but it doesn’t work, unfortunately.
Again with Centos 7 alone everything works. And once the NS installation has completed with the final “system init” game is over.
I think I give up now. But thank you anyway.