Yes, this is it! Thanks Erik! BTW it is quite stupid for me adding by default static route and not using backup router if one will fail, but second is available. In my opinion this "feature with static route" should be disabled by default for IPv6 protocol. 2014/1/21 Erik Auerswald <auerswal@xxxxxxxxxxxxxxxxx>: > Hi, > > I have re-added the LARTC list. Please keep the list in the loop. > > On Tue, Jan 21, 2014 at 11:29:11AM +0100, Vrf Identity wrote: >> Yes I tried it, but without success. >> I also waited for the whole lifetime - entry with lifetime expires >> disappeared but default via fe80::f225:72ff:fed5:74f0 dev eth0 proto >> static metric 1 was still there >> >> notebook:~$ ip -6 rou >> fe80::/64 dev eth0 proto kernel metric 256 >> default via fe80::f225:72ff:fed5:74f0 dev eth0 proto static metric 1 > ^^^^^^ > The static entry will not time out. But you should be able to remove it: > > sudo ip -6 route del default via fe80::f225:72ff:fed5:74f0 dev eth0 proto static metric 1 > >> default via fe80::46e4:d9ff:feeb:f86c dev eth0 proto ra metric 1024 >> expires 1762sec > > The dynamic entry will time out, but the static one will be used anyway if > available. > > Regards, > Erik > >> 2014/1/21 Erik Auerswald <auerswal@xxxxxxxxxxxxxxxxx>: >> > Hi, >> > >> > On Tue, Jan 21, 2014 at 10:28:57AM +0100, Vrf Identity wrote: >> >> I tested IPv6 NDP failover but without success. >> >> >> >> OS: Xubuntu, kernel 3.11.0-15-generic >> >> >> >> >> >> Topology: >> >> >> >> ---- Router 1 >> >> Client - >> >> ---- Router 2 >> >> >> >> >> >> >> >> Automatic IPv6 configuration of network interface. >> > >> > Did you configure the lifetime of the router advertisements sent? >> > The router will be considered available as long as the lifetime has not >> > expired. >> > >> > You can use "ip -6 route" to display the IPv6 routes of the client >> > including the lifetimes. >> > >> >> Code: >> >> notebook:~$ ifconfig >> >> eth0 Link encap:Ethernet HWaddr 00:16:d3:8d:d7:d1 >> >> inet6 addr: fe80::216:d3ff:fe8d:d7d1/64 Scope:Link >> >> inet6 addr: 2011:abc:55:55:216:d3ff:fe8d:d7d1/64 Scope:Global >> >> inet6 addr: 2011:abc:55:55:3861:f333:e0fa:106b/64 Scope:Global >> >> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 >> >> RX packets:57008 errors:0 dropped:31 overruns:0 frame:0 >> >> TX packets:56138 errors:0 dropped:0 overruns:0 carrier:0 >> >> collisions:0 txqueuelen:1000 >> >> RX bytes:6393965 (6.3 MB) TX bytes:6874111 (6.8 MB) >> >> >> >> >> >> Scenario: >> >> >> >> - two gateways with same loopback address >> >> - continuously ping the loopback address 2011:abc:11:11::1 >> >> - after some time disconecting one router >> >> >> >> >> >> After some time client should use second router and connectivity >> >> should back to normal >> >> >> >> >> >> During ping >> >> >> >> Code: >> >> notebook:~$ ip neigh sho >> >> fe80::f225:72ff:fed5:74f0 dev eth0 router REACHABLE >> >> fe80::46e4:d9ff:feeb:f86c dev eth0 lladdr 44:e4:d9:eb:f8:6c router STALE >> >> >> >> After router router fail >> >> >> >> Code: >> >> notebook:~$ ip neigh sho >> >> fe80::f225:72ff:fed5:74f0 dev eth0 router INCOMPLETE >> >> fe80::46e4:d9ff:feeb:f86c dev eth0 lladdr 44:e4:d9:eb:f8:6c router STALE >> >> >> >> First default gateway is still in INCOMPLETE state and client still >> >> want to use it but should change to next. >> >> >> >> Anyone know why? Is is bug in implementation? >> >> Thanks >> > >> > You can read about a similar experiment with Cisco equipment at >> > http://packetlife.net/blog/2011/apr/18/ipv6-neighbor-discovery-high-availability/ >> > >> > HTH, >> > Erik -- To unsubscribe from this list: send the line "unsubscribe lartc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
