Hi guys,
Ever since I upgraded from pfSense 2.1.5 to 2.2, I've had problems with
IPv6 remaining available to all workstations.
My ISP is Internode in Australia, who offers dual stack IPv6 with a 56
bit subnet allocated via DHCPv6 over IPv4. I have my WAN connection set
up as follows:-
IPv6 Configuration Type: DHCP6
Use IPv4 connectivity as parent interface: Checked
Request only a IPv6 prefix: Checked
DHCPv6 Prefix Delegation size: Checked
Send IPv6 prefix hint: Unchecked
The LAN interface is statically set with an IPv6 address and DHCPv6/RA
is set up as follows:-
DHCPv6 server disabled
Router Advertisements: Unmanaged (have also tried Assisted, with a
DHCPv6 server running from my domain controller with the same results)
Router Priority: High
RA Subnet(s): none listed
DNS servers: My domain controller's IPv6 address is listed.
The dashboard shows the following interface information:-
WAN: up
59.167.122.229
fe80::20d:b9ff:fe14:723a
LAN: up
100baseTX <full-duplex>
10.100.18.254
2001:44b8:41d6:ce00::1
When the router first comes up, I have full IPv6 connectivity without an
issue. However, after a couple of hours, IPv6 becomes unavailable from
my Ubuntu 14.04 LTS server box - ping6s go without replies. My Windows
desktop sometimes continues to resolve IPv6 addresses (but can't ping
them), however it often reverts to IPv4, though a ping -6 usually works
fine. My domain controller (Server 2012 Foundation) happily serves up
AAAA records, though often can't ping them. ipv6-test.com shows IPv4
connectivity is fine, but IPv6 is "not supported". DNS6 + IP4 shows up
as reachable, and as I said, my Windows server happily returns AAAA
records if they're found. pfSense happily continues to respond on IPv6
internally, and all communication within my LAN works fine over IPv6.
Servers (of which there are several) have statically configured IPv6
addresses, whereas my desktop configures itself via SLAAC.
A reboot of the router solves the issue, but only for a couple of
hours. The log files aren't much help - routing.log shows a single
entry advising that radvd 1.9.1 started. dhcpd.log shows the following:-
Feb 20 18:54:29 router dhcp6c[15483]: failed to open
/usr/local/etc/dhcp6cctlkey: No such file or directory
Feb 20 18:54:29 router dhcp6c[15483]: failed initialize control message
authentication
Feb 20 18:54:29 router dhcp6c[15483]: skip opening control port
Feb 20 21:02:45 router dhcp6c[15623]: all information to be updated was
canceled
The router was restarted at 18:54. Whilst I can't say for sure that
IPv6 stopped functioning at 21:02, this seems to be about the interval
that IPv6 works. system log shows the system booting up, but nothing
relevant afterwards. Likewise, dmesg.boot shows boot up info, but
nothing else.
Reverting to pfSense 2.1.5 would be somewhat of a pain since I can't
find the backup of my configuration prior to upgrading. I'm fairly sure
I did back up, but it's possible that I neglected to do so. The loss of
IPv6 isn't a killer, it's just a little inconvenient.
Can anyone give me some hints as to where this issue may lie? If I've
missed any relevant information, please let me know. This is driving me
slightly nuts.
_______________________________________________
pfSense mailing list
https://lists.pfsense.org/mailman/listinfo/list
Support the project with Gold! https://pfsense.org/gold
