Still errors e1000

[frankz]

Hello everyone, I don't understand if the error is generated under overload or anything. I noticed that the system signals a hung up of the card, always e1000.

e1000e 0000:00:19.0 enp0s25: Detected Hardware Unit Hang:
                  TDH                  <40>
                  TDT                  <6a>
                  next_to_use          <6a>
                  next_to_clean        <3f>
                buffer_info[next_to_clean]:
                  time_stamp           <10279a12a>
                  next_to_watch        <40>
                  jiffies              <10279a3c0>
                  next_to_watch.status <0>
                MAC Status             <80083>
                PHY Status             <796d>
                PHY 1000BASE-T Status  <3800>
                PHY Extended Status    <3000>
                PCI Status             <10>

The node has 5 network interfaces, one dual with intel chip and one on integrated board.

 

[oguz]

hi,

I noticed that the system signals a hung up of the card, always e1000.

it's a known bug [0] with workaround: ethtool -K <device name> gso off gro off tso off tx off rx off in your case the device is called enp0s25 (you might need to install the ethtool package)

to make it permanent you can add it under the affected interface in /etc/network/interfaces as a post-up directive.

hope this helps

[0]: https://forum.proxmox.com/threads/e1000-driver-hang.58284/

 

[frankz]

hi,


it's a known bug [0] with workaround: ethtool -K <device name> gso off gro off tso off tx off rx off in your case the device is called enp0s25 (you might need to install the ethtool package)

to make it permanent you can add it under the affected interface in /etc/network/interfaces as a post-up directive.

hope this helps

[0]: https://forum.proxmox.com/threads/e1000-driver-hang.58284/

First of all, thank you for answering, it asks you exactly what to put in the interface configuration file and the sequence where to put it. Thank you very much.In addition, having 1 integrated Intel card and the other duals in 3 nodes, is this string valid for the same? I think so as the kernel module is always the e1000. I'm waiting for you

 

[oguz]

First of all, thank you for answering, it asks you exactly what to put in the interface configuration file and the sequence where to put it. Thank you very much.In addition, having 1 integrated Intel card and the other duals in 3 nodes, is this string valid for the same? I think so as the kernel module is always the e1000. I'm waiting for you

the fix is for interfaces using the e1000 driver, you can check your interfaces drivers' with ethtool -i <device name>

 

[frankz]

the fix is for interfaces using the e1000 driver, you can check your interfaces drivers' with ethtool -i <device name>

 

[frankz]

the fix is for interfaces using the e1000 driver, you can check your interfaces drivers' with ethtool -i <device name>

All nodes used the built-in interface and are e1000 as a module

 

[oguz]

All nodes used the built-in interface and are e1000 as a module

then apply the workaround for that interface using its device name and see if the error still occurs.

 

[frankz]

then apply the workaround for that interface using its device name and see if the error still occurs.

Hi, sorry if I misunderstood, I just wanted to ask you how to insert the work around string directly at boot, at momneto I ran via terminal on the node cards, to be clear only on the intel. I don't know how to permanently insert this string on the interface file. Thank you

 

[oguz]

I just wanted to ask you how to insert the work around string directly at boot,

you can add it in your /etc/network/interfaces file, for example:

iface enp0s25 inet manual
        post-up ethtool -K enp0s25 tso off gso off

auto vmbr0
iface vmbr0 inet static
        address 192.168.1.100/20
        gateway 192.168.1.1
        bridge-ports enp0s25
        bridge-stp off
        bridge-fd 0
        post-up ethtool -K enp0s25 tso off gso off