PVE with 2 NICs - unable to connect via SSH/HTTP

D

derTom

New Member
May 16, 2025
3
1
1
Hello,

I have a Lenovo ThinkCentre M720q Tiny PC running Proxmox VE 8.4.1. I have installed an additional NIC for the Fritzbox guest network (IoT devices). Everything works perfectly until I need to restart PVE—as was the case today when there was a power outage.

The server boots correctly and I can ping the internal IP, but the web interface is not accessible. SSH is also not available. Only when I perform a hard shutdown, remove the guest LAN cable, and then boot it up again, am I able to log into the web interface. Once I immediately reconnect the second LAN cable, everything works flawlessly—until the next upgrade/reboot...

Below are the syslog entries:

Code: 
May 16 12:47:49 homie kernel: e1000e 0000:00:1f.6: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
May 16 12:47:49 homie kernel: r8169 0000:02:00.0: enabling device (0000 -> 0003)
May 16 12:47:49 homie kernel: e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
May 16 12:47:49 homie kernel: e1000e: Intel(R) PRO/1000 Network Driver

May 16 12:47:49 homie kernel: r8169 0000:02:00.0 eth0: jumbo features [frames: 9194 bytes, tx checksumming: ko]
May 16 12:47:49 homie kernel: r8169 0000:02:00.0 eth0: RTL8168h/8111h, 00:0a:cd:00:xx:xx, XID 541, IRQ 126

May 16 12:47:49 homie kernel: e1000e 0000:00:1f.6 eth1: MAC: 13, PHY: 12, PBA No: FFFFFF-0FF
May 16 12:47:49 homie kernel: e1000e 0000:00:1f.6 eth1: Intel(R) PRO/1000 Network Connection
May 16 12:47:49 homie kernel: e1000e 0000:00:1f.6 eth1: (PCI Express:2.5GT/s:Width x1) e8:6a:64:fa:xx:xx
May 16 12:47:49 homie kernel: e1000e 0000:00:1f.6 0000:00:1f.6 (uninitialized): registered PHC clock

May 16 12:47:50 homie kernel: e1000e 0000:00:1f.6 eno1: renamed from eth1
May 16 12:47:50 homie kernel: r8169 0000:02:00.0 enp2s0: renamed from eth0

May 16 13:01:49 homie kernel: Gast: port 1(enp2s0) entered forwarding state
May 16 13:01:49 homie kernel: Gast: port 1(enp2s0) entered blocking state
May 16 13:01:49 homie kernel: r8169 0000:02:00.0 enp2s0: Link is Down
May 16 13:01:49 homie kernel: Generic FE-GE Realtek PHY r8169-0-200:00: attached PHY driver (mii_bus:phy_addr=r8169-0-200:00, irq=MAC)
May 16 13:01:49 homie kernel: r8169 0000:02:00.0 enp2s0: entered promiscuous mode
May 16 13:01:49 homie kernel: r8169 0000:02:00.0 enp2s0: entered allmulticast mode
May 16 13:01:49 homie kernel: Gast: port 1(enp2s0) entered disabled state
May 16 13:01:49 homie kernel: Gast: port 1(enp2s0) entered blocking state
May 16 13:01:48 homie kernel: e1000e 0000:00:1f.6 eno1: entered promiscuous mode
May 16 13:01:48 homie kernel: e1000e 0000:00:1f.6 eno1: entered allmulticast mode
May 16 13:01:48 homie kernel: vmbr0: port 1(eno1) entered disabled state
May 16 13:01:48 homie kernel: vmbr0: port 1(eno1) entered blocking state

And here is my /etc/network/interfaces:
Code: 
auto lo
iface lo inet loopback

iface eno1 inet manual

iface enp2s0 inet manual

auto vmbr0
iface vmbr0 inet static
        address 192.168.10.100/24
        gateway 192.168.10.254
        bridge-ports eno1
        bridge-stp off
        bridge-fd 0

auto Gast
iface Gast inet static
        address 192.168.179.100/24
        bridge-ports enp2s0
        bridge-stp off
        bridge-fd 0

Finally the routes:
Code: 
default via 192.168.10.254 dev vmbr0 proto kernel onlink
192.168.10.0/24 dev vmbr0 proto kernel scope link src 192.168.10.100
192.168.179.0/24 dev Gast proto kernel scope link src 192.168.179.100

Any ideas to solve this?

Thanks, Tom
 
Hi @derTom


Possible problem:

Your bridge Gast may attempt to initialize before enp2s0 (Realtek NIC) fully brings the link up, causing the bridge to go into blocked state.

Add pre-up delay to the guest NIC bridge
Allow time for the Realtek NIC to negotiate link before the bridge tries to use it.

auto Gast
iface Gast inet static
address 192.168.179.100/24
bridge-ports enp2s0
bridge-stp off
bridge-fd 0
pre-up sleep 5
Click to expand...

You can increase to sleep 10 if needed.

Let us know if it helped, otherwise there will be option with change drivers for R8169.

I do not have this adapter, so can't simulate.
 
Last edited:
Hi Lukas,
Setting the sleep parameter was a good idea, but unfortunately it was not successful. I even increased it to 15 seconds. Can you help me with your other suggestion with the driver change?
 
so was fine until you added the 2nd NIC? you could switch bridge membership to see if the problem follows the NIC or stays with the bridge
 
Hi Lukas,
sorry for my late answer. I have no possibility to switch to another nic. It's a m2 Slot-NIC. The selection is not very large and the chipset is a plain vanilla one.
I read the given link but the problems mentioned there do not fit mine
Maybe I should live with the problem temporarily and wait for the release of Debian 13 in summer.
 
  • Like
Reactions: Lukas Moravek
You must log in or register to reply here.
Top Bottom