the interface not found message might indicate that your interfaces were renamed with the new kernel - check the output of `ip link`, and update your /etc/network/interfaces - see also:
https://pve.proxmox.com/pve-docs/chapter-sysadmin.html#_naming_conventions
I hope this helps!
@Stoiko - Really appreciate the suggestion to compare ip link - however, it seems the kernel thing might be a red herring....or maybe I'm reading it wrong.
This is really frustrating and I'm super confused.
This is the latest 6.11 kernel on Proxmox:
Code:
# uname -a
Linux grandstand-vm02 6.11.0-2-pve #1 SMP PREEMPT_DYNAMIC PMX 6.11.0-2 (2024-12-04T10:29Z) x86_64 GNU/Linux
And the output of
ip link from the Proxmox host, booted under this 6.11 kernel:
https://gist.github.com/victorhooi/4385bbeff0eb6489c85637a61df9aad4
And this is the latest 6.8 kernel on Proxmox
Code:
# uname -a
Linux grandstand-vm02 6.8.12-6-pve #1 SMP PREEMPT_DYNAMIC PMX 6.8.12-6 (2024-12-19T19:05Z) x86_64 GNU/Linux
And the same output of
ip link from the Proxmox host, booted under this 6.8 kernel:
https://gist.github.com/victorhooi/582ea69a1c8ee9215e45bb805ff590e5
From a quick diff - the device names are the same between the two - just the MAC addresses have all changed - does that seem right?
When I look at the VM layer - when the VMs boot up, they have no network connectivity. And this is the output of ip link from an example VM:
Code:
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: ens18: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN mode DEFAULT group default qlen 1000
link/ether bc:24:11:bd:a1:f8 brd ff:ff:ff:ff:ff:ff
altname enp6s18
altname enxbc2411bda1f8
3: tailscale0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1280 qdisc fq_codel state UNKNOWN mode DEFAULT group default qlen 500
link/none
If I try
sudo ifupdown ens18 it replies with
ifup: unknown interface ens18. However,
sudo ifup end6s18 does work - this successfully gets a DHCP lease.
Another data-point - I have a TrueNAS Scale VM which doesn't seem to hit this issue.
But all the Debian-based OSes have this issue =(.
And there's also an ArubaOS VM also has this issue =(. (AFAIK, ArubaOS is HPE's own custom embedded Linux OS - not sure the lineage).
The last part - where ArubaOS also hits this issue - is why I am thinking it's something that's changed in the Proxmox setup - since the ArubaOS VM hasn't been updated in several months.
Here's also the domes output from a VM:
https://gist.github.com/victorhooi/f66285d0ed3c94e6574f547549e9a2a1
I did see this line in there - but I think that's normal:
Code:
[ 1.100303] virtio_net virtio2 ens18: renamed from eth0
Are there any other obvious things to check here? Have there been any other major changes in Proxmox, or would this still potentially be kernel update related?
