I have two connections: eth0 and ppp0 over it. I've set up quagga and configured zebra and ripd so I can get routes from my ISP (through eth0). It seems to work as expected (
route -n shows lots of routes, local traffic goes through eth0 directly and Internet traffic goes through ppp0) until I switch off VPN connection. After that
route -n shows almost empty table, but when I connect to ripd through telnet (
telnet localhost 2602)
show ip rip status shows ripd received routes less than 30 seconds before and
show ip rip shows lots of routes. But when I connect to zebra through telnet (
telnet localhost 2601) and do
show ip route timestamps only increase (does it mean zebra doesn't get updates from ripd?). Restarting ripd solves the problem.
I'm using Fedora 11 with KDE, configured VPN connection through KNetworkManager applet.
/etc/quagga/zebra.conf
Code:
hostname NNNNN
password XXX
enable password XXX
log file /var/log/quagga/zebra.log
/etc/quagga/ripd.conf
Code:
router rip
version 2
network eth0
passive-interface default
line vty
log file /var/log/quagga/ripd.log
password XXX
When I turn off VPN connection the following messages appear in ripd log:
Code:
2009/06/19 17:30:21 RIP: interface delete ppp0 index 7 flags 4240 metric 0 mtu 1
2009/06/19 17:30:22 RIP: setsockopt_multicast_ipv4 attempting to drop and re-add (fd 5, ifaddr 10.55.88.116, mcast 224.0.0.9, ifindex 2)
Could anyone please help me with this?