r8168/9 disconnect after PVE8 upgrade

D

dannybridi

New Member
Sep 10, 2022
12
0
1
Hi everyone,

I've seen many with similar issues, but no resolution that I could apply (at least I wasn't able to).

I have a test PC that was running fine with PVE7. After the upgrade to PVE8, the network connection would stop after a few hours after reboot. The screen shows loads of messages like:
[xxxxxx.xxxxxx] r8169 0000:01:00.0 enp1s0: rtl_ephyar_cond == 1 (loop: 100, delay: 10)

All I could do is reboot to reset the connection.

lspci shows that I have a Realtek RTL8111/8168/8411:

root@pve3:~# lspci -nnvmm | egrep -A 6 -B 1 -i 'network|ethernet'
Slot: 01:00.0
Class: Ethernet controller [0200]
Vendor: Realtek Semiconductor Co., Ltd. [10ec]
Device: RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [8168]
SVendor: Dell [1028]
SDevice: RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller [080c]
Rev: 15
ProgIf: 00


I first re-installed the server (from a PVE8 ISO) thinking that the upgrade went wrong, but that did not help.

I followed some threads where I installed the r8168-dkms package, but all that did was completely disconnect the server from the network. So, I re-installed again (I'm not a Linux expert :) ).

Any help in resolving this issue would be appreciated.

Thanks
 
Hello
I also have exactly the same problem.
Disabling "weak by LAN" in the BIOS helped for a while - the server worked much longer, but it still crashed every few days.
An IT specialist advised me this method (version B):

https://community.hetzner.com/tutorials/installing-the-r8168-driver

...however, something went wrong and Proxmox disabled the network card completely. Despite trying, I was unable to reconnect :/
Currently, I managed to install HA without virtualization and I'm testing... but I'd also like to know how to easily solve the problem with "r8169"
Regards
 
Thanks @Tikos.
Strangely enough, the server has not shown any issues since the last re-install, 2.5 days ago! The difference, I should have mentioned before, is that after the last install, I not join this node to a cluster. So, it’s a standalone node. I will wait another day or so before joining it to a cluster. I wonder if that makes any difference.
 
Update: as mentioned above, I left the server non-clustered for a week without any issues. I joined it again yesterday to a cluster (a total of 3 nodes). Sure enough, the server lost network connectivity some time after that. This seems to indicate that the issue is also related to clustering.
 
Last edited:
I haven't had this problem since I installed Home Assistant without Proxmox. However, this still does not solve the problem with Proxmox.
 
Having the same r8169 issue. Non-clustered node. Remoting into MS Server2019 guest with remote desktop appears to trigger it. So far connecting through QEMU hasn't.

Update: It just failed in QEMU as well
 
Last edited:
Same problem here with r8169 and PVE 8.0.

randomly loosing connection, and installing r8168-dkms package doesn't solve the issue, as dannybridi said, it breaks the connection entirely :)
 
dannybridi said:
Update: as mentioned above, I left the server non-clustered for a week without any issues. I joined it again yesterday to a cluster (a total of 3 nodes). Sure enough, the server lost network connectivity some time after that. This seems to indicate that the issue is also related to clustering.
Click to expand...
I'm seeing the issue in non-clustered nodes. Clustering was so unstable in our 20 node environment that we took the opportunity to break it while doing the 7 to 8 upgrade.
 
We only saw this issue on our 6 Dell OptiPlex 3060. At this moment 2 of these are on the r8169 driver and have yet to drop the nic. The remaining 4 are on the r8168 driver. Unfortunately, I discovered today that 2 of these were offline. The other 2 have been solid for almost 2 weeks. The 2 that were offline were also showing an ACPI error. These are the steps I used to roll back to the r8168 driver. Apologizes to the original poster for not crediting them. I can't find their website:



add to /etc/apt/sources.list:

deb http://ftp.debian.org/debian bookworm main contrib non-free non-free-firmware
deb http://ftp.debian.org/debian bookworm-updates main contrib non-free non-free-firmware

apt update
apt dist-upgrade
apt install pve-headers
apt install r8168-dkms
echo blacklist r8169 >> /etc/modprobe.d/blacklist-r8169.conf
ethtool -i enp1s0
 
  • Like
Reactions: dgallant
I don't want to claim victory yet but I have solved my problem (at least for now) by disabling the power options in my BIOS:

photo_2023-10-12_01-11-47.jpg
 
Hello. I'm having same problems as you all, after some minutes to hours working I cannot connect to the PVE instance or VM's anymore. So I decided to downgrade from PVE 8 to PVE 7 with original r8169 and no efforts on any command line repairing anything and now everything is working now. It is a pity that version 8 is buggy of my DELL MFF 3000 Series i5 9500t. Anyway I stay tuned at this forum conversation when someone really find the
 
jcopeland said:
We only saw this issue on our 6 Dell OptiPlex 3060. At this moment 2 of these are on the r8169 driver and have yet to drop the nic. The remaining 4 are on the r8168 driver. Unfortunately, I discovered today that 2 of these were offline. The other 2 have been solid for almost 2 weeks. The 2 that were offline were also showing an ACPI error. These are the steps I used to roll back to the r8168 driver. Apologizes to the original poster for not crediting them. I can't find their website:



add to /etc/apt/sources.list:

deb http://ftp.debian.org/debian bookworm main contrib non-free non-free-firmware
deb http://ftp.debian.org/debian bookworm-updates main contrib non-free non-free-firmware

apt update
apt dist-upgrade
apt install pve-headers
apt install r8168-dkms
echo blacklist r8169 >> /etc/modprobe.d/blacklist-r8169.conf
ethtool -i enp1s0
Click to expand...
Thank you for the rundown.

For folks like me, that installed fresh and without a subscription, you have to add/edit sources.list:

# Proxmox VE pve-no-subscription repository provided by proxmox.com,
# NOT recommended for production use
deb http://download.proxmox.com/debian/pve bookworm pve-no-subscription

Otherwise the
Code: 
apt install pve-headers
will not work!
 
Last edited:
here's what ours look like. had to add the non-free-ware repo's for the pve-headers to install

/etc/apt/sources.list
deb http://ftp.us.debian.org/debian bookworm main contrib
deb http://ftp.debian.org/debian bookworm main contrib

deb http://ftp.us.debian.org/debian bookworm-updates main contrib
deb http://ftp.debian.org/debian bookworm-updates main contrib

# security updates
deb http://security.debian.org bookworm-security main contrib
deb http://security.debian.org/debian-security bookworm-security main contrib

# not for production
deb http://download.proxmox.com/debian bookworm pve-no-subscription
deb http://download.proxmox.com/debian/pve bookworm pve-no-subscription

#non-free-firmware
deb http://ftp.debian.org/debian bookworm main contrib non-free non-free-firmware
deb http://ftp.debian.org/debian bookworm-updates main contrib non-free non-free-firmware


/etc/apt/sources.list.d/pve-enterprise.list
#deb https://enterprise.proxmox.com/debian/pve bookworm pve-enterprise


/etc/apt/sources.list.d/ceph.list
#deb https://enterprise.proxmox.com/debian/ceph-quincy bookworm enterprise
 
You must log in or register to reply here.
Top Bottom