On further investigation, it appears this problem has something to do
with the state of the network driver... despite netstat -g showing
that eth0 is in the correct multicast group, the multicast RIP packets
are never being accepted, hence never getting to the ripd. This can be
remedied in a workaround by doing 'ifconfig eth0 allmulti" or
"ifconfig eth0 promisc"-- there is also something (?) that tcpdump,
without its -p option, does which seems to resolve the issue, but
neither PROMISC nor ALLMULTI show up in ifconfig. This lasts (as far
as I can tell) until the next reboot.
I chose to add the following to the /etc/network/interfaces entry for
eth0:
post-up /sbin/ifconfig eth0 allmulti
just in case it helps anyone seeing the same problem, here are some
relevant dmesg lines from my driver:
natsemi dp8381x driver, version 1.07+LK1.0.17, Sep 27, 2002
natsemi eth0: NatSemi DP8381[56] at 0xd0832000 (0000:00:0e.0),
00:a0:cc:a1:a5:13, IRQ 10, port TP.
I'm using kernel-image-2.6.8-2-686 2.6.8-16sarge1
If I'm the only one seeing this ripd behavior, it's probably OK to
close the bug and leave this info for anyone with similar problems
looking into a similar problem.
thanks for all the help, folks.
- M
--
Mark "Monty" Montague | [EMAIL PROTECTED] | I don't do Windows(tm)
If a tree falls when there's no one observing, does its wave function collapse?
<URL:http://www.gg.caltech.edu/~monty/monty.shtml>
X-GNUPG-Fingerprint: E1E3 E513 3687 94E3 5C61 C97B CBC7 2CA6 01B1 F0A7
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
