Everytime I restart my incus host none of these containers receive an iP, If I restart the container or I restart systemd-networkd they do receive it. I have checked all revelant logs for systemd-networkd that I know of and I still cannot determine whats going on. I started doing some troubleshooting inside the container logs as follows:
root@Python:~# dmesg
dmesg: read kernel buffer failed: Operation not permitted
journalctl shows a bunch of errors like the following:
Apr 17 00:42:14 Python udevadm[136]: event_source: Failed to write ‘add’ to ‘/sys/bus/event_source/uevent’: Permission denied
Apr 17 00:42:14 Python udevadm[136]: gpio: Failed to write ‘add’ to ‘/sys/bus/gpio/uevent’: Permission denied
Apr 17 00:42:14 Python udevadm[136]: hardware_error_device: Failed to write ‘add’ to ‘/sys/bus/acpi/drivers/hardware_error_device/uevent’: Permission denied
Apr 17 00:50:44 MyDesktop udevadm[137]: LNXSYSTM:00: Failed to write ‘add’ to ‘/sys/devices/LNXSYSTM:00/uevent’: Permission denied
Apr 17 00:50:44 MyDesktop udevadm[137]: LNXPWRBN:00: Failed to write ‘add’ to ‘/sys/devices/LNXSYSTM:00/LNXPWRBN:00/uevent’: Permission denied
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8201F Fast Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8201F Fast Ethernet/uevent’: Permissio>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8208 Fast Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8208 Fast Ethernet/uevent’: Permission >
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8211 Gigabit Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8211 Gigabit Ethernet/uevent’: Permi>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8211B Gigabit Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8211B Gigabit Ethernet/uevent’: Per>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8211C Gigabit Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8211C Gigabit Ethernet/uevent’: Per>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8211DN Gigabit Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8211DN Gigabit Ethernet/uevent’: P>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8211E Gigabit Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8211E Gigabit Ethernet/uevent’: Per>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8211F Gigabit Ethernet: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8211F Gigabit Ethernet/uevent’: Per>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8211F-VD Gigabit Ethernet: Failed to write ‘add’ to '/sys/bus/mdio_bus/drivers/RTL8211F-VD Gigabit Ethernet/uevent>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8221B-VB-CG 2.5Gbps PHY: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8221B-VB-CG 2.5Gbps PHY/uevent’: P>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8221B-VM-CG 2.5Gbps PHY: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8221B-VM-CG 2.5Gbps PHY/uevent’: P>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8226 2.5Gbps PHY: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8226 2.5Gbps PHY/uevent’: Permission deni>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8226-CG 2.5Gbps PHY: Failed to write ‘add’ to ‘/sys/bus/mdio_bus/drivers/RTL8226-CG 2.5Gbps PHY/uevent’: Permissio>
Apr 17 00:58:28 MyDesktop udevadm[133]: RTL8226B-CG_RTL8221B-CG 2.5Gbps PHY: Failed to write ‘add’ to '/sys/bus/mdio_bus/drivers/RTL
if I do a incus restart all I see all seems to work and all containers get either their respective static ip or dynamic ip. This is only happening when I restart the host.
mainbr0 is externally managed at the os using bridge-utils and the interfaces file config below. Also I keep wondering if all the permission denied messages are part of the issue or not related. Below is my network config for the host.
root@HomeServer:~# brctl show mainbr0
bridge name bridge id STP enabled interfaces
mainbr0 8000.72ddbd7207de no enp1s0
vethba6083a7
vethbef6a858
vethd7134385
I did further research, I assigned a static MAC to the bridge using the primary network card. That did not seem to help. I went inside the container and did systemctl enable systemd-networkd that seems to have done the trick I now see IP`s on all containers both static and dynamic. Does the could image not use systemd-networkd by default? Anywhere I can read up on this? I still do see the errors I mentioned and the following when I check eth0 in an instance.
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: Failed to increase receive buffer size for general netlink socket, ignoring: Operation not permitted
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: eth0: Link UP
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: eth0: Gained carrier
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: lo: Link UP
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: lo: Gained carrier
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: eth0: Gained IPv6LL
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: Enumeration completed
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: eth0: Configuring with /etc/systemd/network/10-cloud-init-eth0.network.
Apr 17 03:00:38 MyDesktop systemd[1]: Started systemd-networkd.service - Network Configuration.
Apr 17 03:00:38 MyDesktop systemd-networkd[154]: eth0: DHCPv4 address 10.10.0.106/8, gateway 10.10.0.1 acquired from 10.10.0.1
now that I crosscheck my previous post, could it be something is disabling systemd-networkd upon reboot? Maybe the permissions issue?
ok something is definately maybe wrong with my config. I create a new instance, only with the default profile, did not change networking or any type of configuration.
incus init images:debian/12/cloud MyDesktop
then went inside did journalctl -b and see a ton of permission errors.
The sysctl and module errors are normal, you’re in a container, containers can’t load modules or modify system-wide sysctls, that’s fine.
Why are you using the cloud variant of the image? Do you actually use cloud-init?
I’ve not looked at the Debian cloud images specifically but it wouldn’t surprise me if cloud-init on Debian was still using ifupdown which would then explain why systemd-networkd is disabled.
I was using cloud-init to setup my user, and installed applications. Not totally required though.i am choosing Debian/12 cloud images as I was thinking theyre the most stable and I am familiar with them. Any recommendations?
The base debian/12 image should be basically the same but without cloud-init which I believe also means that it will be using networkd out of the box for networking rather than whatever cloud-init does on Debian.
Hi Just an update so I installed the debian/12 images as instructed yesterday. I just booted up the pc and to my surprise once again my systemd-networkd is disabled any other ideas? If I restart networkd or I reboot the client it works, but not upon initial host boot. I do notice one thing, one of my other containers has docker installed, which does seem to work upon boot. I read on the docs that docker can cause problems while running on the host. On the host I only run incus. Does it matter if the host is running ENI etc/network/interfaces and the containers run systems?
root@Bashing:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
6: eth0@if7: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 00:16:3e:55:81:4a brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet6 fe80::216:3eff:fe55:814a/64 scope link
valid_lft forever preferred_lft forever
root@Bashing:~# networkctl
WARNING: systemd-networkd is not running, output will be incomplete.
IDX LINK TYPE OPERATIONAL SETUP
1 lo loopback - unmanaged
6 eth0 ether - unmanaged
2 links listed.
Errors from journal on systemd