Upgraded to Proxmox VE 9 - Intel i219-V nic down network wont work

Z

zenowl77

Active Member
Feb 22, 2024
244
45
28
Okay so i upgraded from VE 8.4 to 9 and everything seemed to have gone fine but after a reboot, the nic wont work.

It was down so i ran “ip link set dev enp0s31f6 up” and now it is up with no ip. (And of course doesn’t stay up after a reboot)

Any suggestions/ideas on how i can fix this problem?

Thank you.
 
Maybe during the upgrade your nic was renamed to enp0s31f6
Check your network config at /etc/network/interfaces and replace the old interface name with enp0s31f6

hope this helps
 
that was the name prior to the upgrade thankfully so no changes needed for /etc/network/interfaces i also double checked the file just to be sure.
 
The interface name seems quite long, is this a multi port nic?
If it has multiple ports maybe the ports got renamed?
 
Also share ip a. I like to use dhclient -v to find the right port.
 
Last edited:
  • Like
Reactions: PwrBank
It is in fact a single port nic and after trying to get it started a few more times it seems to have bound to the wrong ip and isnt using vmbr0 like it is supposed to either so i cannot access webui but i can access ssh so thats a start i guess to help fix it…

/etc/network/interfaces:
Code: 
auto lo
iface lo inet loopback

iface enp0s31f6 inet manual
        mtu 1500

auto vmbr0
iface vmbr0 inet static
        address 10.0.0.23/24
        gateway 10.0.0.1
        bridge-ports enp0s31f6
        bridge-stp off
        bridge-fd 0
        bridge-vlan-aware yes
        bridge-vids 2-4094
        mtu 1500
        pre-up ethtool -K enp0s31f6 tx off gso off <—- required on 8.4 to function
        post-up ethtool -K vmbr0 tx off gso off
 
auto vmbr1
iface vmbr1 inet manual
        bridge-ports none
        bridge-stp off
        bridge-fd 0
        bridge-vlan-aware yes
        bridge-vids 2-4094
 
Last edited:
Weird, if you can SSH into it, then it should be working. Is it possible the webui service is crashed or off?
 
PwrBank said:
Weird, if you can SSH into it, then it should be working. Is it possible the webui service is crashed or off?
Click to expand...
yeah, apparently it seems to have just removed half of proxmox and quit, saying it finished so everything was just marked as removed and not actually upgraded so i tried telling it to upgrade again and it finished again seemingly all normal and fine and now its causing a kernel panic at reboot and didn't seem to actually install everything. so not sure why or how but the upgrade just gutted the system and said everything finished normally...
 
My test cluster did the same thing. During the upgrade it was failing to upgrade some type of Python3 application and then saying "all good to go!" and half of the packages were missing.

I ended up re-installing.

See if
Bash: 
apt --fix-broken install
Shows anything. When I ran that, a TON of entries showed up. When it went to install them, it would show which package was stopping the upgrade. However, idk if it's fixable, I tried for a few hours but gave up. I probably shouldn't have install PVE9 on the test cluster anyway because it's using Linstor for storage, and I doubt it's compatible.
 
it looks like i will also just have to completely reinstall, sadly this is also my main server as i only have the one and don't really have the money for multiple or tons of backups, etc, so it just nuked everything......

on mine nothing said it failed at all, everything said it was good to go like it went perfectly smooth, my nvidia drivers even built via DKMS perfectly without error...
 
Oh yeah, the PVE8 to 9 upgrade didn't actually error out, I only knew there was problem because VMs wouldn't boot - Checked the "apt --fix-broken install" command and saw a ton of packages that were half broken, tried to have it repair them - Then it would fail that process with "python3-something" was not building or had some type of error.
 
Everyone has a test environment, but not everyone is lucky enough to have it not in prod. :p
 
You may be able to boot into rescue mode (from pve install iso) and try to copy your vm's and containers to an external usb drive manually ...
 
PwrBank said:
Everyone has a test environment, but not everyone is lucky enough to have it not in prod. :p
Click to expand...
haha unfortunately.


MarkusKo said:
You may be able to boot into rescue mode (from pve install iso) and try to copy your vm's and containers to an external usb drive manually ...
Click to expand...
i think i will likely attempt something like this, i have a few of them backed up already thankfully. but i dont have enough free space to make full backups of all of them sadly.

anyone know if i can reinstall to local and not local-lvm, etc or is there a repair option or anything? its been so long since i installed proxmox, setting up the boot disks now to start working on it, was just going to backup /etc/ and everything else i hav added and restore it after reinstall
 
You must log in or register to reply here.
Top Bottom