[SOLVED] Intel NIC e1000e hardware unit hang

Same here. I ran an update yesterday. But I suspect an earlier PVE update already triggered this already as a guest VM on that NUC runs one of my DNS servers and it seems that one is preferred by many clients. As a result, I notice all kinds of networking issues on my clients which in the end are DNS issues, that are fixed after a while. So, I suspect I have had this issue for a while now. The issue became much worse after the last PVE update until the point that the system became unreachable and I had to go to the console. I have now added the post-up, and configured clients on the network only to use the other DNS for now. But this is such an issue that if not resolved I have to drop the whole setup.
 
gctwnl said:
Same here. I ran an update yesterday. But I suspect an earlier PVE update already triggered this already as a guest VM on that NUC runs one of my DNS servers and it seems that one is preferred by many clients. As a result, I notice all kinds of networking issues on my clients which in the end are DNS issues, that are fixed after a while. So, I suspect I have had this issue for a while now. The issue became much worse after the last PVE update until the point that the system became unreachable and I had to go to the console. I have now added the post-up, and configured clients on the network only to use the other DNS for now. But this is such an issue that if not resolved I have to drop the whole setup.
Click to expand...
Aside from the fact that the root cause is still present after more than three years, you should reconsider your DNS setup. With DNS failover configured, the issue shouldn't affect any clients.
 
TErxleben said:
Aside from the fact that the root cause is still present after more than three years, you should reconsider your DNS setup. With DNS failover configured, the issue shouldn't affect any clients.
Click to expand...
How would DNS failover help me with clients on the local network getting their DNS server config from DHCP? The only thing I can think of is DNS in a HA setup, a la HAproxy, but the latter only does TCP/HTTP, not UDP.

My issue is that I have two DNS servers running on my LAN with two different LAN IP addresses. I have to provide both to the clients (phones, computers) with DHCP. If the clients are stuck on trying to use the one they have been using so far even if it is unreachable, that client is stuck. What I would need is a HA way to run my (unbound) resolvers.
 
gctwnl said:
How would DNS failover help me with clients on the local network getting their DNS server config from DHCP? The only thing I can think of is DNS in a HA setup, a la HAproxy, but the latter only does TCP/HTTP, not UDP.

My issue is that I have two DNS servers running on my LAN with two different LAN IP addresses. I have to provide both to the clients (phones, computers) with DHCP. If the clients are stuck on trying to use the one they have been using so far even if it is unreachable, that client is stuck. What I would need is a HA way to run my (unbound) resolvers.
Click to expand...
The two IP addresses are exactly your problem. This can be easily solved by "hiding" multiple DHCP/DNS servers behind one IP address.

First, search the internet for something like "pihole failover." You'll find plenty of suggestions there.

HA is by no means necessary.
 
TErxleben said:
The two IP addresses are exactly your problem. This can be easily solved by "hiding" multiple DHCP/DNS servers behind one IP address.

First, search the internet for something like "pihole failover." You'll find plenty of suggestions there.

HA is by no means necessary.
Click to expand...
Yes I understand that. I am looking into setting up a virtual IP on my OPNsense router. Creates a SPOF, but if the router dies nothing much works anyway.
 
gctwnl said:
Yes I understand that. I am looking into setting up a virtual IP on my OPNsense router. Creates a SPOF, but if the router dies nothing much works anyway.
Click to expand...
You'll never get rid of SPOFs. However, these problems can be easily reduced to a minimum. DHCP/DNS is quite easy to handle. In my opinion, DHCP/DNS is the most important service in a network.
 
had the driver bug happen to me for the first time a few days ago after updating to updating proxmox to 8.4.1 on my I217-V NIC, my I350 had no issues.

Currently testing after running "ethtool -K eno1 tso off gso off"

have also added to /etc/network/interfaces

iface eno1 inet manual
post-up /sbin/ethtool -K eno1 tso off gso off
#Onboard NIC


lspci -v | grep Ethernet
root@pve:~# lspci -v | grep Ethernet
00:19.0 Ethernet controller: Intel Corporation Ethernet Connection I217-V
Subsystem: Gigabyte Technology Co., Ltd Ethernet Connection I217-V

01:00.0 Ethernet controller: Intel Corporation I350 Gigabit Network Connection (rev 01)
Subsystem: Hewlett-Packard Company Ethernet 1Gb 2-port 361T Adapter

01:00.1 Ethernet controller: Intel Corporation I350 Gigabit Network Connection (rev 01)
Subsystem: Hewlett-Packard Company Ethernet 1Gb 2-port 361T Adapter
 
Hi.
Same error for me.
Now I put this "fix" today and it's working at the moment:

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

iface enp0s25 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 enp0s25 tso off gso off

source /etc/network/interfaces.d/*

This is my ethernet:
Code: 
# lspci -v | grep Ethernet
00:19.0 Ethernet controller: Intel Corporation Ethernet Connection I217-LM (rev 05)
        Subsystem: Dell Ethernet Connection I217-LM

This did not happen with version 8.4. Last Thursday I upgraded to 8.4.1 and it started to fail from April 20.
 
You must log in or register to reply here.
Top Bottom