TX underrun –- Connection Drop

[depster]

For roughly the past 2 weeks, I have been seeing our internet connection drop all connections for a minute or so a few times per day. I am currently corralating it with the following error messages in the system log. How can I correct this?

Jan 18 16:20:50 check_reload_status: rc.linkup starting
Jan 18 16:20:49 dnsmasq[553]: exiting on receipt of SIGTERM
Jan 18 16:20:49 kernel: dc0: link state changed to UP
Jan 18 16:20:47 kernel: dc0: link state changed to DOWN
Jan 18 15:22:07 kernel: dc0: TX underrun – increasing TX threshold
Jan 18 15:14:28 check_reload_status: rc.linkup starting
Jan 18 15:14:24 kernel: dc0: link state changed to UP
Jan 18 15:14:23 kernel: dc0: watchdog timeout
Jan 18 15:14:19 kernel: dc0: link state changed to DOWN
Jan 18 15:13:47 kernel: dc0: link state changed to UP
Jan 18 15:13:25 kernel: dc0: link state changed to DOWN
Jan 18 15:09:43 kernel: dc0: TX underrun -- increasing TX threshold

[jimp]

I have also seen this on dc-based NICs, they seem to just be a really, really, low-quality card. Switching to another brand of card is probably your best bet.