I have a Centos 8 guest that is having network issues. I'm not sure if this is a Centos 8 issue or a Centos 8 on Proxmox issue. But my almost identical Centos 7 guest is not having this issue and has almost an identical configuration (different public IP). The network adapter is ens18 and there is a configuration script in /etc/sysconfig/network-scripts. It is set with onboot=yes. If I use nmcli connection up ens18, it works fine. But after a reboot, the network connection no longer works. If I go into nmtui, it will show two instances of ens18, one with a gateway and one without. If I remove the instance without a gateway and reboot, the extra instance appears again. If I remove the instance with the gateway and reboot, it stays with a single instance but doesn't work. And nmcli connection up ens18 doesn't work, even after setting the gateway. I have to remove all instances and re-add it manually, after which I can get it up again, until a reboot. Rince. Repeat. Any ideas? I'm at a complete loss here. The only thing somewhat unique is that the gateway is on a different subnet as they are public IPs provided by OVH. Not sure if that could be part of the issue.
[SOLVED] Centos 8 Network Boot Issue
- Thread starter bionemesis
- Start date
hi,
i'd check what is being run when vm boots. maybe somewhere in the network you have a setting, which tries to put the interface up?
this seems suspicious. if your guest is a virtual machine, proxmox doesn't do any modifications for the network config (since it's distro based it all happens inside the vm). so most likely something is misconfigured in the vm.
i'd check what is being run when vm boots. maybe somewhere in the network you have a setting, which tries to put the interface up?
Well I don't have all the details yet, but it's related to something that ApisCP does during installation. It may be related to https://access.redhat.com/solutions/4973191.