[SOLVED] Network dead after update (8.1.10)

[Noobie]

Hi all!

After updating my Proxmox (pve-manager/8.1.10/4b06efb5db453f29 (running kernel: 6.5.13-5-pve), my network connection no longer works (lights remain dark). Unfortunately, I can no longer access the GUI. I am using a USB LAN adapter from Ugreen.

The USB device seems to be found. Output from lsusb:

Bus 002 Device 002: ID 0b95:1790 ASIX Electronics Corp. AX8819 Gigabit Ethernet

The output of ip a looks like this, where eth0 is DOWN:


When I do and then ifup vmbr0 (after running ifdown vmbr0) I get:


My /etc/network/interfaces looks like this (I haven't changed anything!):



Unfortunately I don't know how to fix this. Do you have an idea?

 

[WindowLock]

What does the command ip link show show you?
It looks like the "iface" routing is wrong after your upgrade.

As I can find online (perhaps a search direction) https://superuser.com/questions/1000570/why-eth0-is-renamed-to-enx78e7d1ea46da-after-upgrade

 

[Iodok]

What is the output of:
ethtool eth0 ?

 

[gfngfn256]

Unfortunately I don't know how to fix this. Do you have an idea?

Your output of ip a above appears to show an eth0 device - do you have another onboard ethernet port - apart from the USB adapter?

If you don't - then it appears that your old USB adapter name enx000.... has been changed to eth0. You'll need to change the /etc/network/interfaces accordingly.

If you do have another onboard ethernet port - then it appears your USB adapter is not currently being picked up by the network management of OS.

 

[Noobie]

Thanks for answering!

This solved my problem:

Your output of ip a above appears to show an eth0 device - do you have another onboard ethernet port - apart from the USB adapter?

If you don't - then it appears that your old USB adapter name enx000.... has been changed to eth0. You'll need to change the /etc/network/interfaces accordingly.

If you do have another onboard ethernet port - then it appears your USB adapter is not currently being picked up by the network management of OS.

Thanks for your help!

 

[rschaskos]

Hello everyone, I have a similar problem.

After updating my Proxmox (pve-manager/8.2.4/faa83925c9641325 (running kernel: 6.8.8-1-pve), my network connection no longer works (lights remain dark). Unfortunately, I can no longer access the GUI. I am using a USB LAN adapter from Ugreen.

pveversion:



lsusb:



ip a:



ifup vmbr0:




cat /etc/network/interfaces:



What do you recommend doing to adjust this?

Thanks

 

[gfngfn256]

What do you recommend doing to adjust this?

From the second line of your lsusb output, it appears that a gigabit ethernet USB device is detected.
On your ip a output you show (what appears to be) 2 physical NIC ethernet ports. One is the enx000ec684408f (the old one used in your config) and one named enp9s0. If you do not have / know what that second NIC (enp9s0) is, then most likely your physical enx000ec684408f has been renamed to enp9s0. It would be however interesting that both are shown.

What I would do, is firstly completely shutdown & power off the server (remove power cable/connector, switch plug etc.) so that NO power is going to any of the computer parts. Then power on the server again & see what output you get.

If this does not resolve the situation, & you cannot identify another NIC on your server, then change the /etc/network/interfaces file; replacing both references of enx000ec684408f to enp9s0 .

However in your situation, I suspect there really is another NIC, and something is up with the new kernel (6.8) concerning your USB adapter.
If that proves the case, you will have to revert to the previous working kernel on your server.

 

[rschaskos]

From the second line of your lsusb output, it appears that a gigabit ethernet USB device is detected.
On your ip a output you show (what appears to be) 2 physical NIC ethernet ports. One is the enx000ec684408f (the old one used in your config) and one named enp9s0. If you do not have / know what that second NIC (enp9s0) is, then most likely your physical enx000ec684408f has been renamed to enp9s0. It would be however interesting that both are shown.

What I would do, is firstly completely shutdown & power off the server (remove power cable/connector, switch plug etc.) so that NO power is going to any of the computer parts. Then power on the server again & see what output you get.

If this does not resolve the situation, & you cannot identify another NIC on your server, then change the /etc/network/interfaces file; replacing both references of enx000ec684408f to enp9s0 .

However in your situation, I suspect there really is another NIC, and something is up with the new kernel (6.8) concerning your USB adapter.
If that proves the case, you will have to revert to the previous working kernel on your server.

Thanks for your reply.

Exactly.
I'm using an old dell inspiron 1545 laptop as a server. As it doesn't have a gigabit ethernet connection, I decided to add a UGREEN USB Ethernet adapter.

Therefore, the NIC enp9s0 refers to the laptop's physical connection and the NIC enx000ec684408f refers to the UGREEN Ethernet connection.

Since I installed proxmox on this laptop I have never used the enp9s0 NIC, only the USB Ethernet. In other words, I didn't change these settings before or after the proxmox update.

If you need any other information, just ask and I will send it.

Do you think I should change the kernel version back?

 

[gfngfn256]

Do you think I should change the kernel version back?

I assume you already followed my idea of fully powering down the server (as above) - but this did not help.

There has been quite a few problems reported with the current kernel - especially with USB devices (3.0?) - & therefore for the time being I would pin your kernel to a previous working one.

 

[rschaskos]

I assume you already followed my idea of fully powering down the server (as above) - but this did not help.

There has been quite a few problems reported with the current kernel - especially with USB devices (3.0?) - & therefore for the time being I would pin your kernel to a previous working one.

Thanks for the feedback.

Yes! I've turned the device off, remove all connector, power cable, and on again, but it didn't help.
How could I return my kernel to a previous version? I'm not a Linux expert... do I need to put version 7 again?
I have many active containers, such as DNS, pfSense, Torrent, Plex, Home Assistant, among others.

Remembering that I cannot access via GUI.

Thank you.

 

[gfngfn256]

do I need to put version 7 again

No. What you need to do is pin to an older kernel, as described here.

Start by showing output for proxmox-boot-tool kernel list

 

[rschaskos]

No. What you need to do is pin to an older kernel, as described here.

Start by showing output for proxmox-boot-tool kernel list

Hello

Let me show the output for command

 

[mr.c]

I assume you already followed my idea of fully powering down the server (as above) - but this did not help.

There has been quite a few problems reported with the current kernel - especially with USB devices (3.0?) - & therefore for the time being I would pin your kernel to a previous working one.

That worked for me, using the proxmox-boot-tool

 

[rschaskos]

That worked for me, using the proxmox-boot-tool

Tell me if my understanding is correct:

Check available kernels: proxmox-boot-tool kernel list

In my case I have two versions:
5.15.158-1-pve (previous)
6.8.8-1-pve (current)

Fix previous version: proxmox-boot-tool kernel pin 5.15.158-1-pve --next-boot

Update settings: proxmox-boot-tool refresh

 

[gfngfn256]

No - it appears you updated from Proxmox 7.4 directly to the current 8.2 version. Maybe what I would do in your case is download the proxmox-ve_8.1-2.iso & reinstall. You can find a list of all available ISOs here.

MAKE SURE YOU HAVE ALL YOUR VMs & LXCs BACKUPs ON AN EXTERNAL MEDIUM
ALSO HAVE ALL NOTES ABOUT YOUR INSTALL HOST ETC. HANDY

 

[rschaskos]

No - it appears you updated from Proxmox 7.4 directly to the current 8.2 version. Maybe what I would do in your case is download the proxmox-ve_8.1-2.iso & reinstall. You can find a list of all available ISOs here.

MAKE SURE YOU HAVE ALL YOUR VMs & LXCs BACKUPs ON AN EXTERNAL MEDIUM
ALSO HAVE ALL NOTES ABOUT YOUR INSTALL HOST ETC. HANDY

What would be the best way to perform a restorable backup, as I do not have access to the GUI.

 

[gfngfn256]

What would be the best way to perform a restorable backup, as I do not have access to the GUI.

IDK what you have setup as Storage Backup. If you show the output for cat /etc/pve/storage.cfg we can have a look.

 

[gfngfn256]

So the command for doing backups is vzdump
You can see its CLI options/usages here.

 

[rschaskos]

Tell me if my understanding is correct:

Check available kernels: proxmox-boot-tool kernel list

In my case I have two versions:
5.15.158-1-pve (previous)
6.8.8-1-pve (current)

Fix previous version: proxmox-boot-tool kernel pin 5.15.158-1-pve --next-boot

Update settings: proxmox-boot-tool refresh

Hey guys, I managed to solve the problem with these commands above:

But I used the command without --next-boot: proxmox-boot-tool kernel pin 5.15.158-1-pve

As you said dear gfngfn256, the problem was the kernel version.

Thank you all

best regards