ipv6 only failover wan network interface stops routing
I have a firewalla gold plus running 1.981 with two wan links:
- Quantum which is IPv4 only as primary. ipv6 is disabled
- Xfinity which is IPv4 and v6 as failover, ipv6 is enabled
Immediately after restarting my device I am able to connect externally via any device on the network to v6 sites:
mkortekaas@marks ~> ping6 google.com
PING6(56=40+8+8 bytes) 2601:282:1c02:5861:cdd0:eb9b:d798:c8f1 --> 2607:f8b0:400f:802::200e
16 bytes from 2607:f8b0:400f:802::200e, icmp_seq=0 hlim=115 time=13.205 ms
16 bytes from 2607:f8b0:400f:802::200e, icmp_seq=1 hlim=115 time=10.966 ms
16 bytes from 2607:f8b0:400f:802::200e, icmp_seq=2 hlim=115 time=10.540 ms
16 bytes from 2607:f8b0:400f:802::200e, icmp_seq=3 hlim=115 time=12.521 ms
However, after "a while" ipv6 stops working fully:
mkortekaas@marks ~> ping6 google.com
PING6(56=40+8+8 bytes) 2601:282:1c02:5861:cdd0:eb9b:d798:c8f1 --> 2607:f8b0:400f:802::200e
^C
--- google.com ping6 statistics ---
19 packets transmitted, 0 packets received, 100.0% packet loss
Rebooting the firewalla resolves this issue. Also if I change a device to route all prefer the xfinity network for all traffic v6 works perfectly
-
Correct quantum is the primary and I don't see any outages in the interface since the prior reboot. This seems to happen after a "few days" but I have just created an hourly check script which will check and pull information from the firewalla and will update when I next detect the issue
-
It took 15 days for this to appear again - in looking at the logs the xfinity interface went down just before the issue occurred - looking at the router table as I see:
pi@Firewalla:~ (XXXXXXX) $ ip -6 neigh show dev eth0
fe80::21c:73ff:fe00:99 lladdr 00:1c:73:00:00:99 router STALEI then tried:
pi@Firewalla:~ (XXXXXXX) $ sudo ip link set eth0 down
sleep 60
sudo ip link set eth0 up
pi@Firewalla:~ (XXXXXXX) $ ip -6 neigh show dev eth0
fe80::226d:31ff:fe31:bebb lladdr 20:6d:31:31:be:bb STALE
fe80::21c:73ff:fe00:99 lladdr 00:1c:73:00:00:99 router STALEAfter a reboot the issue resolved itself. Not sure why a down/up would not do the trick on this but I couldn't seem to find any other command that would reset that stale issue. After reboot:
pi@Firewalla:~ (XXXXXXX) $ ip -6 neigh show dev eth0
2001:558:6040:a2:15d6:8914:7508:93af FAILED
fe80::226d:31ff:fe31:bebb lladdr 20:6d:31:31:be:bb STALE
fe80::21c:73ff:fe00:99 lladdr 00:1c:73:00:00:99 router DELAY
Please sign in to leave a comment.
Comments
4 comments