Hi folks,
I am running a carp environment on my gateway. Due to lack
of routable IPv4 addresses the em0 interface provides IPv6
only, the carp0 interface defines both IPv4 and IPv6 addresses.
The internal interfaces em1 and carp1 provide both IPv4 and
IPv6.
ntpd works fine on the master, but on the backup host ntpd
complains "---- peer not valid ----" and "sendto: Network is
unreachable". "ntpctl -s peers" shows *no* IPv6 addresses
(on master and backup), even though there seems to be some
IPv6 support in the code.
The workaround is clear, but I wonder how comes?
ntpd.conf:
# grep -v ^\# /etc/ntpd.conf
listen on 10.100.0.1
listen on 10.100.0.3
listen on 2001:db8:30:fff0::1
listen on 2001:db8:30:fff0::3
servers pool.ntp.org
servers ntp.eu.sixxs.net
The packet filter allows ntp as well:
pass out log quick proto udp from (self) to any port ntp keep state
(no-sync)
There is no nat-to for (self).
Every helpful comment is highly appreciated.
Harri
