[SOLVED] e100e hang issue

[companion7102]

Edit: I searched and found different topics but I didn't saw that there waas a huge thread with 90+ replies and no solutions due to firmware issue.
If this post needs to be deleted, no problem.

Hello,

Like other I have a hang on my ethernet that happens at some random time but sure thing, the connection can't pass a full day.

Here is the error from the System log of my pve01:

Jun 27 08:30:28 pve01 kernel: e1000e 0000:00:1f.6 eno1: Detected Hardware Unit Hang:
  TDH                  <a4>
  TDT                  <c7>
  next_to_use          <c7>
  next_to_clean        <a3>
buffer_info[next_to_clean]:
  time_stamp           <100f8ed09>
  next_to_watch        <a4>
  jiffies              <102ac5640>
  next_to_watch.status <0>
MAC Status             <40080083>
PHY Status             <796d>
PHY 1000BASE-T Status  <3800>
PHY Extended Status    <3000>
PCI Status             <10>

My host is a Lenovo m720q i5-9500T with proxmox installed on it (pve01), no other proxmox instance.
In PVE, I have 1 VM (vm-docker) that is a debian running Docker in it.

Here are some details:
Kernel: Linux 6.8.12-11-pve (2025-05-22T09:39Z)
Manager version: pve-manager/8.4.1

From pve01 shell

lspci | grep -Ei "ethernet"
00:1f.6 Ethernet controller: Intel Corporation Ethernet Connection (7) I219-V (rev 10)


ip link show | grep -A2 -E "eno1|vmbr0"
2: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master vmbr0 state UP mode DEFAULT group default qlen 1000
    link/ether e8:6a:64:9e:dc:8a brd ff:ff:ff:ff:ff:ff
    altname enp0s31f6
--
4: vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000
    link/ether e8:6a:64:9e:dc:8a brd ff:ff:ff:ff:ff:ff
5: tap100i0: <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master fwbr100i0 state UNKNOWN mode DEFAULT group default qlen 1000
--
7: fwpr100p0@fwln100i0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master vmbr0 state UP mode DEFAULT group default qlen 1000
    link/ether be:9e:39:96:29:39 brd ff:ff:ff:ff:ff:ff
8: fwln100i0@fwpr100p0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master fwbr100i0 state UP mode DEFAULT group default qlen 1000

I've read other posts here and tested what is often suggested on this particular post: https://forum.proxmox.com/threads/e1000-driver-hang.58284/page-8#post-390709

I first tested to switch the VM (vm-docker) from VirtIO (paravirtualized) to Intel E1000E.
The next morning the vm, containers and PVE weren't accessible and I force stopped it.
I switched back the VM to VirtIO (paravirtualized).

Next, I tried in pve01 to disable the offloads as suggested. First on vmbr0 because it is the one with the IP conf, I tought it made sens.
Later that day, same hand result. So I did the same on eno1 after a forced reboot of the host.

Note that before editing the interfaces file, I entered the command
ethtool -K vmbr0 gso off gro off tso off tx off rx off rxvlan off txvlan off sg off
ethtool -K eno1 gso off gro off tso off tx off rx off rxvlan off txvlan off sg off

Here is my pve01 /etc/network/interfaces

auto lo
iface lo inet loopback

iface eno1 inet manual
        offload-gso off
        offload-gro off
        offload-tso off
        offload-rx off
        offload-tx off
        offload-rxvlan off
        offload-txvlan off
        offload-sg off
        offload-ufo off
        offload-lro off

auto vmbr0
iface vmbr0 inet static
        offload-gso off
        offload-gro off
        offload-tso off
        offload-rx off
        offload-tx off
        offload-rxvlan off
        offload-txvlan off
        offload-sg off
        offload-ufo off
        offload-lro off
        address 192.168.1.7/24
        gateway 192.168.1.1
        bridge-ports eno1
        bridge-stp off
        bridge-fd 0

iface wlp2s0 inet manual

source /etc/network/interfaces.d/*

I rebooted the host (pve01) after each modification of /etc/network/interfaces.

vm-docker /etc/network/interfaces

source /etc/network/interfaces.d/*

# The loopback network interface
auto lo
iface lo inet loopback

# The primary network interface
allow-hotplug ens18
iface ens18 inet dhcp
# This is an autoconfigured IPv6 interface
iface ens18 inet6 auto

I don't know what I can do or test next, any help will be greatly appreciated. Thank you.

 

[MarkusKo]

i think it does not work because just putting offload-* into your network config does nothing.
these are params for ethtool so you have to do it like that:

# cat /etc/network/interfaces
auto lo
iface lo inet loopback

iface eno1 inet manual

auto vmbr0
iface vmbr0 inet static
        address 192.168.188.94/24
        gateway 192.168.188.1
        bridge-ports enp0s25
        bridge-stp off
        bridge-fd 0
        post-up /sbin/ethtool -K eno1 tso off gso off

source /etc/network/interfaces.d/*

source: https://forum.proxmox.com/threads/intel-nic-e1000e-hardware-unit-hang.106001/post-765582

the post-up hook invokes ethtool which does set the offload settings

 

[companion7102]

Thanks for the post-up, I missed that yes.
I tried and I still have hangs at some point. Maybe my eth card is not working properly at this point. It's not a brand new host, used one.
I switched to an USB3 1GBps adapter ASIX AX88179. Looks to work properly at the moment.