Q: A RIPng neigbour doesn't seem to get the other's updates
A: Juan: Probably there's some connectivity failure, the best form of testing RIPng connectivy is trying to send a ping like this:
# ping6 ff02::9%fxp0or
# ping6 -I fxp0 ff02::9ff02::9 is the link-local router multicast address, you should get a reply for each one of your routers who are connected to the same link. If you run tcpdump, you should see packets like these:
11:43:25.871282 fe80::2c0:26ff:fea3:68f4.521 > ff02::9.521: ripng-resp 3: 2001:720:410:100a::/64 (1)[|ripng]This example shows that RIPng uses port 512 as well.
Q: How can I pick up traffic towards a fake direction ?
A: Juan: This is useful to collect packets towards NATing routers. This is achieved by adding "route" clause into "router ripng" section:
route 2001:720:410:1001:1::/96
Q: I have a mixed enviroment with Linux/Cisco/FreeBSD RIPng routers. Linux routers don't get the default route announced by other routers' "default-information originate" clause.
A: Juan: There was a bug in linux IPv6 routing table (probaply already fixed). The best form to cope with this is to annouce a route like this:
route 2000::/3and to forget the "default-information originate" command.
Q: I've got more than 16 hops, and RIPng seems not to be working
A: Juan: Use OSPF for IPv6.