Hello all,

 

I have eth0 set for no bond no bridge. This interface used for the Unraid access / shares and plex docker.  Eth1 is setup on its own vlan on a different network with bridging enabled. I receive this in the logs and it is messing with ip address and access to unraid.  

 

ar 29 04:07:24 Unraid kernel: vethcbcc6ea: renamed from eth0

Mar 29 04:07:25 Unraid kernel: eth0: renamed from veth9f1932f

Mar 29 04:07:30 Unraid kernel: veth3005291: renamed from eth0

Mar 29 04:07:57 Unraid kernel: eth0: renamed from vethe239e55

Mar 29 04:08:02 Unraid kernel: vethf978886: renamed from eth0

Mar 29 04:08:14 Unraid kernel: eth0: renamed from veth811c7d3

Mar 29 04:08:21 Unraid kernel: vetha7ab26d: renamed from eth0

Mar 29 04:08:23 Unraid kernel: eth0: renamed from veth4bd73f7

Mar 29 04:08:34 Unraid kernel: vethfda06d3: renamed from eth0

Mar 29 04:08:43 Unraid kernel: eth0: renamed from veth5908af7

Mar 29 04:08:45 Unraid kernel: veth5e20ec6: renamed from eth0

Mar 29 04:08:45 Unraid kernel: eth0: renamed from veth0a7cf4d

 

My docker is set to ipvlan not mac.

 

 

Docker custom network type 

ipvlan

 

Host access to custom networks:

Disabled

 

Preserve user defined networks:

No

 

IPv4 custom network on interface br1:

Subnet: 192.168.4.0/24

 

With all the latest changes to the docker network setting in recent releases I am not sure what is causing this. I do have a ubiquity throughout my home if that matters.

 

When this first stated I decided to install an intel nic just for the majority of my docker containers. I move all of them off of eth0 except for plex. etho is the motherboards nic.  The intel e1000 nic is a 2 port nic and I could disable eth0 and move it to eth2 if that would help stability. 

 

When eth0 goes down and I cannot access Unraid I can still access the other dockers on eth1 on the other network. So, I know the box is not locked up just the eth0.

 

Thank you for any light you canb shine on this issue.  

I get the same issue here, the odd thing is, I use Ubiquiti as well, and I have the server plugged right into the switch.

 

When I loose access, I can see the switch having no lights blinking.  If you disconnect the "ethernet" cable from the switch nada.

If you disconnect the ethernet cable from the back of the server/tower, bham, it lights up and everything back to normal.

 

Same pattern, such as:

Jun 27 09:00:19 Tower kernel: veth6b32b85: renamed from eth0
Jun 27 09:00:19 Tower kernel: br-986610407cef: port 1(veth9b6cc1b) entered disabled state
Jun 27 09:00:20 Tower kernel: device veth9b6cc1b left promiscuous mode
Jun 27 09:00:23 Tower kernel: eth0: renamed from veth576eedf
Jun 27 09:00:23 Tower kernel: IPv6: ADDRCONF(NETDEV_CHANGE): vetha27e76b: link becomes ready
Jun 27 09:00:23 Tower kernel: br-986610407cef: port 1(vetha27e76b) entered blocking state
Jun 27 09:00:23 Tower kernel: br-986610407cef: port 1(vetha27e76b) entered forwarding state  
  

Unsure if related to the drivers, of the NIC.

ASUSTeK COMPUTER INC. TUF GAMING X570-PRO (WI-FI) , Version Rev X.0x American Megatrends Inc., Version 5003

AMD Chipset.

And uses the Intel igcIntel(R) 2.5G Ethernet Linux DriverInusenet/ethernet/intel/igc

Ethernet controller: Intel Corporation Ethernet Controller I225-V (rev 03)

 

I am on the latest public release for Unraid, but the nic might be lagging.

- igc driver and firmware are up version is 6.1.79-Unraid, and the firmware version is 1057:8754. 

 

I do see a Firmware as of 2 months ago, so I'll give that a shot, I'd check on that as well and see if it fixes your issue, and I'll report back on mine.

 

Edited June 27, 20242 yr by RoTalk
typos

Did you find any solution? I am having same issues.

the eth0 rename by itself is normal, it's when a container starts, if that keeps happening there's likely a container always re-starting, if there are other issues that may not be the cause.

On 10/7/2024 at 11:55 PM, JorgeB said:

there's likely a container always re-starting

What is the easiest method to determine which docker container is frequently restarting?

Check the up times for all containers.