Subject: mc0: underflow
To: None <port-macppc@netbsd.org>
From: Daniel Lamblin <daniell@trillian.beachbash.net>
List: port-macppc
Date: 02/10/2003 22:05:59
This is the built in ethernet interface on an 8500.
When I issued: pkg_add apache, all I got was a message indicating that
there was no inet6 route to the host (ftp.netbsd.org).
Then the interface went down.  I left it to think, seeing if it would
recover, but eventually I had to log into the console and issue an
ifconfig for the interface again, though ifconfig said the device was up
before, it only started working again and responding to ssh sessions once
I ran ifconfig.

Obviously I don't like haivng to resort to the console at random.
Has anyone else encountered such problems?
This didn't happen the next time I ran pkg_add, nor again.  Maybe I should
disable indet6, since my hadware doesn't really support it.
This machine is DMZed "behind" a linksys dsl router/switch.

Here's a log of messages:
Feb  8 07:00:10 tnt syslogd: restartFeb  8 13:59:58 tnt /netbsd: mc0: underflow
Feb  8 14:06:56 tnt ntpd[161]: sendto(216.231.41.2): Network is down
Feb  8 14:08:57 tnt ntpd[161]: sendto(216.254.95.2): Network is down
Feb  8 14:14:33 tnt ntpd[161]: sendto(128.59.59.177): Network is down
Feb  8 21:55:57 tnt ntpd[161]: Connection re-established to 128.59.59.177
Feb  8 22:05:09 tnt ntpd[161]: Connection re-established to 216.231.41.2
Feb  8 22:07:05 tnt ntpd[161]: Connection re-established to 216.254.95.2
Feb  8 22:11:26 tnt ntpd[161]: time reset 0.405766 s
Feb  9 00:00:01 tnt syslogd: restart