[SOLVED] Network Devices Renamed after apt upgrade & power outage - Can't connect to the network

S

sdsys

New Member
Jul 1, 2026
3
1
3
Hi folks, any help would be amazing here.
I had a power outage this past weekend and ever since my PVE server has been unable to get an IP address on my local network. I also noticed that instead of the usual 4 network devices, I only have 3. I'm concerned that something may have shocked a malfunction into a network card.
I've so far tried the following
- rebooting
- ifreload -a
- updating /etc/network/interfaces to match the new names of the network devices (tried all 3 available network devices but none of them worked)
- Removing the IP Reservation on my router that previously existed for this device, then forcing /etc/network/interfaces to use DHCP (no response)
- Removing/Reseating/Replacing network cables (switch doesn't recognized the cables as plugged in - no activity lights on the port)
- Replacing the switch the computer's plugged into (IPMI Interface connects, but server's network cards don't.
- Plugging server directly into router as oppoosed to switch (switch is a TP-Link Festa FS328G, router is a TP-Link Deco M5)
- Updating firmware on Motherboard as well as IPMI interface (mobo is an ASRock RACK EP2C621D12 WS)
Up next is to try an ethernet dongle in one of the USB ports, see if I can get an IP that way. I need to buy one of those though

Attached is contents of /etc/network/interfaces
20260701_121940.jpg

ip a
20260701_121918.jpg

and /etc/hosts
20260701_123439.jpg


All the conventional wisdom I found from searching hasn't been able to help me. Additionally I've never seen device names like this, nor have I seen one of the 4 ethernet ports on the motherboard just disappear like this. I can't find anything with the prefix tap10x anywhere

If anyone's got ideas please share them. Thank you for reading all this and I hope you have a lovely day
 
- Dont't use apt upgrade
- Why is /etc/interfaces formatted so poorly? No identation at all
- Why is your host commented out in /etc/hosts?
- tap* is most certainly not the right interface but I see no real one. Please share lspci -nnk
- Did you use PCI(e) passthrough?
- What's journalctl -krp warning say?
 
Last edited:
  • Like
Reactions: sdsys
Hi,
Checking dmesg or lspci would reveal whether the hardware is still physically present on the PCIe bus. Because it seems to be missing
 
  • Like
Reactions: sdsys
Impact said:
- Dont't use apt upgrade
- Why is /etc/interfaces formatted so poorly? No identation at all
- Why is your host commented out in /etc/hosts?
- tap* is most certainly not the right interface but I see no real one. Please share lspci -nnk
- Did you use PCI(e) passthrough?
- What's journalctl -krp warning say?
Click to expand...
Please excuse the poor formatting. Last time this happened I was able to fix it by just updating the interface names in /etc/interfaces.
The host was commented out in /etc/hosts as a testing method. That's when I still thought it was a DHCP problem

lspci -nnk output below
20260701_134900.jpg
No PCI Passthrough involved here, the ethernet devices are baked onto the mobo

journalctl -krp warning output below
20260701_134830.jpg

Apologies for the pics, if I could ssh in and take regular screenshots I would.


fba said:
Hi,
Checking dmesg or lspci would reveal whether the hardware is still physically present on the PCIe bus. Because it seems to be missing
Click to expand...
lspci shows the devices are present, and dmesg didn't show anything promising besides what's in the pics above.

Thank you both for your replies
 
sdsys said:
No PCI Passthrough involved here, the ethernet devices are baked onto the mobo
Click to expand...
Just trying to rule out that you blacklisted the drivers or similar.

I'd say this is the important part to concentrate on
1782931046733.png
Unfortunately I'm not quite sure why that would happen or how to fix it. What's supposed to help is to disconnect power for a bit?
There's also a memory error. Maybe running memtest86 (included on PVE installer iso) might be an idea.
 
Last edited:
  • Like
Reactions: sdsys
Funny thing, I guess I had learned that previously and just forgot - my update script is apt update && apt full-upgrade lol. I am not just an idiot, but a forgetful one, too!
 
  • Like
Reactions: sdsys
Impact said:
Just trying to rule out that you blacklisted the drivers or similar.

I'd say this is the important part to concentrate on
View attachment 98755
Unfortunately I'm not quite sure why that would happen or how to fix it. What's supposed to help is to disconnect power for a bit?
There's also a memory error. Maybe running memtest86 (included on PVE installer iso) might be an idea.
Click to expand...
I'm a bit flabbergasted. Unplugged the server and shut off the power supply itself, waited about 15 minutes, tried again - devices all have appropriate names. I undid the changes I made in /etc/network/interfaces and rebooted, and we are back online.

Thank you to everyone who's put forth their wisdom and knowledge on this!
 
  • Like
Reactions: packetmonkey
You must log in or register to reply here.
Top Bottom
Back