[786674.738193] e1000e 0000:84:00.0 enp132s0f0: Detected Hardware Unit Hang:
                  TDH                  <56>
                  TDT                  <aa>
                  next_to_use          <aa>
                  next_to_clean        <55>
                buffer_info[next_to_clean]:
                  time_stamp           <12edf758f>
                  next_to_watch        <56>
                  jiffies              <12edf9140>
                  next_to_watch.status <0>
                MAC Status             <80387>
                PHY Status             <792d>
                PHY 1000BASE-T Status  <3c00>
                PHY Extended Status    <3000>
                PCI Status             <10>
[786674.929889] e1000e 0000:84:00.0 enp132s0f0: NETDEV WATCHDOG: CPU: 30: transmit queue 0 timed out 7180 ms
[786674.947330] e1000e 0000:84:00.0 enp132s0f0: Reset adapter unexpectedly

Solution

The solution looks to be disabling two features (TSO (TCP Segmentation Offload) or GSO (Generic Segmentation Offload)) using ethtool.

ethtool -K enp132s0f0 tso off gso off

You can add this to your /etc/network/interfaces as a post-up command to maintain this over reboots.

References

• https://forum.proxmox.com/threads/trap-error-on-e1000-network-adapter.105758/
• https://forum.proxmox.com/threads/e1000-driver-hang.58284/page-4#post-302307
• https://docs.kernel.org/networking/segmentation-offloads.html

Categories: TECHNICAL

shlee