Proxmox 6.8.12-9-pve kernel has introduced a problem with e1000e Driver and network connection lost after some hours

It will increase cpu usage and slows the network a bit due to all the smaller packets. But it's a workaround for now I guess, I rolled back to 8.2.2 with Linux 6.8.4-2-pve kernel. USB-Ethernet adapter was just too slow. This isn't just a proxmox issue, on Fedora desktop my intel nic also drops out all the time.
 
Last edited:
Disabling offloading takes a bit of a toll on small boxes with weaker CPUs, I dropped to 400-600Mbps from a full gigabit. Intel I219-LM

Upgraded to PVE 9 (6.14.8-2), re-enabled offloading and hoped for the best, the first few hours have been good so far (this will age like milk).

EDIT: as expected, the good times didn't last long, took about 48 hours before i hit it again
 
Last edited:
i am testing this:
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
Didn't help. Proxmox uses an EOL kernel 6.14.8-11, maybe the problem was fixed in 6.16 but I didn't wait because the network card was failing several times a day. Bought and installed tp-link tx201 (rtl8125) the problem went away without offloadings, everything works stably
 
Yep same issue ! I have lost my connection internet for the past 3 days every 12h my stuff crash...
First I thought it could be the e1000e driver after checking logs I have reduced it to InterruptThrottleRate=3000 setting but still lost internet...
Would this be the solution:
Code: 
ethtool  -K enp0s31f6 gso off gro off tso off tx off rx off rxvlan off txvlan off sg off
thank you
Edit: This command
Code: 
ethtool  -K enp0s31f6 gso off gro off tso off tx off rx off rxvlan off txvlan off sg off
This command unfortunately fail to work on the new kernel update... Only solution will to downgrade kernel I guess. Will keep updating!
Edit only downgrading kernel fixed the issue !
 
Last edited:
Hello, good morning. I arrived here via a Google search.
I have a Dell 7070 with Intel e1000e and these problems.
Today I did an upgrade and there was a new kernel, 6.14.8-3-bpo12-pve.
With this kernel, after two minutes, the network connection is lost.
Is there a solution to this problem?
 
After searching around, I found this recommendation:
Create the file (if it doesn't exist) /etc/modprobe.d/e1000e.conf
Put the following inside:

options e1000e InterruptThrottleRate=3000

Then run:
update-initramfs -u -k all
And reboot.

At the moment, it doesn't cut out after 2 minutes like it did with this kernel. We will keep you informed.
 
  • Like
Reactions: InsanityFlea
I first tried "the hack" directly against the eno1 device section, and performance was terrible.... 226Mbit/sec transfers with the VMs.

I now did this under vmbr0, and performance is back to gigabit speeds. I am yet to see with time if it will hold-up or go offline again.

eno1 is the phisical NIC. Any thoughts? This is my interfaces file:

Code: 
auto lo
iface lo inet loopback

iface eno1 inet manual
#    post-up ethtool -K eno1 tso off gso off

auto vmbr0
iface vmbr0 inet static
    address 10.1.1.9/24
    gateway 10.1.1.1
    bridge-ports eno1
    bridge-stp off
    bridge-fd 0
    post-up ethtool -K eno1 tso off gso off

source /etc/network/interfaces.d/*
 
Just a tip, you only need tso off.
There's no need to turn off all others.
I've been using this on /etc/network/interfaces for over 2 years with no issues:

Code: 
post-up ethtool -K eno1 tso off
Code: 
post-up ethtool -K vmbr0 tso off
 
  • Like
Reactions: Hegse and MarkusKo
I'm also experiencing this on an HP 800 g6 mini running Proxmox 9 .0.10 which has an Intel Corporation Ethernet Connection (11) I219-LM based NIC.

Here is the message that appears in the syslog.

Code: 
Sep 30 08:18:08 pve-02 kernel: e1000e 0000:00:1f.6 eno1: Detected Hardware Unit Hang:
  TDH                  <22>
  TDT                  <d6>
  next_to_use          <d6>
  next_to_clean        <21>
buffer_info[next_to_clean]:
  time_stamp           <1133ba6b9>
  next_to_watch        <22>
  jiffies              <1133d0f40>
  next_to_watch.status <0>
MAC Status             <80083>
PHY Status             <796d>
PHY 1000BASE-T Status  <3800>
PHY Extended Status    <3000>
PCI Status             <10>

I'll disable tso and see how that goes.
 
Darkbotic said:
Just a tip, you only need tso off.
There's no need to turn off all others.
I've been using this on /etc/network/interfaces for over 2 years with no issues:

Code: 
post-up ethtool -K eno1 tso off
Code: 
post-up ethtool -K vmbr0 tso off
Click to expand...
And for those of us who have run the proxmenu script for this card (which I understand disables everything), what should we do?

How do we enable everything and just leave this one off?
 
Darkbotic said:
Just a tip, you only need tso off.
There's no need to turn off all others.
Click to expand...
Hi

Why is only this one deactivation sufficient? There are many others who say it works with TX off and RX off. Is there any reason for this?

Regards, Arp
 
Darkbotic said:
Just a tip, you only need tso off.
There's no need to turn off all others.
I've been using this on /etc/network/interfaces for over 2 years with no issues:

Code: 
post-up ethtool -K eno1 tso off
Code: 
post-up ethtool -K vmbr0 tso off
Click to expand...
Thanks, just to be clear just add
Code: 
post-up ethtool -K eno1 tso off
Code: 
post-up ethtool -K vmbr0 tso off
to /etc/network/interfaces adn reboot?
 
You must log in or register to reply here.
Top Bottom