e1000e network issue on Proxmox

The ethtool configuration is about disabling HW offloading (or acceleration if you will) of certain features. Turning them off just moves the responsibility to software instead (which Linux have done for ages). I'm not aware that this changes anything related to data integrity.
 
Hello, I'm facing the same issue, and I don't know why but the post-up command did nothing.
So I asked Claude.ai to write me a script to simulate a physical unplug replug of the cable.

This is what it did.
Can anyone have a look and criticize ? :D
what do you think ?

thx a lot
 

Attachments

I applied this to my interface file, reboot since, and had the problem yesterday, so it seems it does not work for me :/
 
here it is

Code: 
iface eno1 inet manual
        post-up ethtool -K eno1 gso off tso off rxvlan off txvlan off gro off tx off rx off sg off
 
I created a simple and easy fix for this problem.

Simply disable every troublesome offload feature once at boot with systemd:

1. Check and use active interface: ip a

2. create systemd file and replace enp0s31f6 with the active interface from ip a:
vim /etc/systemd/system/disable-e1000e-offload.service
insert the snippet below
______________________________________________________________________________________________________________
[Unit]
Description=Disable all offloads (incl. VLAN) on e1000e
After=network-online.target
Wants=network-online.target

[Service]
Type=oneshot
ExecStart=/usr/sbin/ethtool -K enp0s31f6 tso off gso off gro off rx off tx off rxvlan off txvlan off
RemainAfterExit=yes

[Install]
WantedBy=multi-user.target
______________________________________________________________________________________________________________

4. Manage service:
systemctl daemon-reload
systemctl enable --now disable-e1000e-offload.service"

5. Verify - alternative reboot server and check again:
ethtool -k enp0s31f6 | grep -E 'segmentation|checksumming|vlan'
 
This is not a valid solution for me, sure, the network does not hang with the offloads turned off, but it drops to FE (100Mbit)
Anyone else seen this?
 
Pete_79 said:
This is not a valid solution for me, sure, the network does not hang with the offloads turned off, but it drops to FE (100Mbit)
Anyone else seen this?
Click to expand...
If you tried the service solution i provided in #29 and rebooted the server then please provide the output from the command below? I have a feeling this is a router/switch issue.

ethtool enp0s31f6 | grep -E 'Speed|Duplex|Link detected'
 
Last edited:
msvarton said:
Can you please provide the output from the command below? I have a feeling this is a router/switch issue.

ethtool enp0s31f6 | grep -E 'Speed|Duplex|Link detected'
Click to expand...
I recently rebooted the host machine, so now it shows 1000Mb/s

root@node3:~# ethtool eno1 | grep -E 'Speed|Duplex|Link detected'
Speed: 1000Mb/s
Duplex: Full
Link detected: yes
root@node3:~#

I can post again as soon as it drops down to 100Mb/s
 
You must log in or register to reply here.
Top Bottom