Proxmox Kernel Update 6.8.4-2 network stops working

I have the: Supermicro X11SPH-nCTF

I updated and restarted as per proxmox suggestion as there was a new kernel installed, 6.8.4-2. Upon reboot, thing came back up. IMPI in, check network, and links are down.
I think the names were changed is part of the issue. Rolled back to Kernel 6.5.13-5 and everything works again.

2: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq master vmbr0 state UP group default qlen 1000
    link/ether 3c:ec:ef:xx:xx:xx brd ff:ff:ff:ff:ff:ff
    altname enp103s0f0
3: eno2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether 3c:ec:ef:xx:xx:xx brd ff:ff:ff:ff:ff:ff
    altname enp103s0f1

I forgot to write down the new names, but they were not enoX something longer named.

Other strange part was, both said state DOWN, even though my switch sayes link negotiated fine at 10g, moved to 1G port still negoitated fine via switch config, but proxmox says down.

How can i fix this? wait for Proxmox to update? if that’s just what will happen with the newer kernel, could someone please guide me on how to fix once I reboot!?

Thank you!

Does this help?

Thanks, that appears to be the issue I am facing… However, no clear solution. someone wrote some kind of script, that i’m not clear if it would really work.
Anyone here with the same motherboard have the issues and fixed?

I used the intructions here Overriding network device names and updated all the interface names to “eth#” and then did the updates to 8.2 and had no network issues.

1 Like

Thanks. I guess i have some work to do to get the upgrade… want to keep down time minimal as one can’t make these changes while the system in online unfortuantly.
This really should be fixed by proxmox :frowning:

My understanding is the issue with the interface names is coming from systemd which is part of the Debian base they use, though I believe and don’t quote me on it that they are looking to change how they handle interfaces in a future release to help mitigate these issues. I have been using Proxmox since the 6.X version and this is only 1 of a few update that I would have had issues with had I not done the interface changes before hand, The other update that caused me issues was around PCI-e passthrough and I believe was an upstream problem they had to wait to get a fix for but was not too long before it was back to working soomthly again.

1 Like

You should still have the older kernel installed, boot from that and pin it using proxmox-boot-tool kernel pin kernel_version for a quick fix.

Proxmox team definitely should have been ahead of this one.

yes, that’s how i got it back up and running for now. I my try the “pin interface name” script/approach, or wait for a fix.