Bug 24978
| Summary: | "dc" network interface goes into continuous reset | ||
|---|---|---|---|
| Product: | Base System | Reporter: | howard <howard> |
| Component: | kern | Assignee: | freebsd-bugs (Nobody) <bugs> |
| Status: | Closed FIXED | ||
| Severity: | Affects Only Me | ||
| Priority: | Normal | ||
| Version: | 4.1.1-RELEASE | ||
| Hardware: | Any | ||
| OS: | Any | ||
This may be a bogus report. The computer showed a major hardware fault
when rebooted (kernel panic - trap 12 - failure to page in) and the
problem with the dc0 interface might very well be an indication that
the driver is doing just what it should - reacting gracefully to an
out-of-the-ordinary condition. I couldn't let you know sooner since
the machine with the problem is our mail relay....
Sorry for the probable false alarm.
>>Walt
State Changed From-To: open->closed probably hardware problem. |
After some period of time (90 days in the latest case, but sometimes less than one day), one of the interface cards begins reporting "watchdog timeout" at roughly one-second intervals. The driver code suggests that this message is associated with resetting the entire interface, and it cannot be pinged from other computers while in this state. The problem seemed worse with earlier releases of FreeBSD. This does not panic the kernel or have a noticeable effect on other processes. Fix: Replace network interface with a different kind that takes a different driver? I like the Tulip interfaces; they and their driver provide more interesting information in my system logs when anything odd is on my network, than do other interfaces. How-To-Repeat: Boot up computer. Watch system log.