The kernel-image package was upgraded from linux-image-5.4.0-169-generic to linux-image-5.4.0-176-generic, so the problem appears to have occurred within that range of kernel versions.
We worked around the problem by configuring grub to boot the "Linux 5.4.0-169-generic" kernel. The affected servers have been rock solid running that kernel but were unusable running the 5.4.0-176-generic kernel.
We encountered similar errors after applying updates to Ubuntu 20.04 LTS in mid-April, 2024:
```
bnxt_en 0000:c8:00.0 eno1np0: TX timeout detected, starting reset task!
```
The kernel-image package was upgraded from linux-image- 5.4.0-169- generic to linux-image- 5.4.0-176- generic, so the problem appears to have occurred within that range of kernel versions.
We worked around the problem by configuring grub to boot the "Linux 5.4.0-169-generic" kernel. The affected servers have been rock solid running that kernel but were unusable running the 5.4.0-176-generic kernel.
/rjg