Hi.
I try on Debian 11/nftables v0.9.8 (E.D.S.) following setup for ipv6,
nftrace being activated.
Traffic coming on wireguard IF fo IP 2001:xxxx::5870/128 should go out
through the wireguard IF, other traffic using default gw.
main and default IF=enp35s0
wireguard if=wig0
On incoming packet, in table mangle
chain INPUT {
type filter hook input priority mangle; policy accept;
meta mark set ct mark
mark != 0x0 accept
iif "wig0" ip6 daddr { 2001:67c:1298::/48 } meta mark set 0x100 ct
mark set meta mark
}
following nftrace I see that incoming packets are going till
trace id dc77c61b ip6 filter INPUT packet: iif "wig0" ip6 saddr
2001:yyyy::24 ip6 daddr 2001:xxxx::5870 ip6 dscp cs0 ip6 ecn not-ect ip6
hoplimit
6 ip6 flowlabel 729556 ip6 length 24 icmpv6 type echo-request icmpv6
code no-route icmpv6 parameter-problem 772505885 @th,64,96 0
trace id dc77c61b ip6 filter INPUT rule meta mark 0x00000100 accept
(verdict accept)
First problem, a tshark capture shows [Frame is marked: False]:
shouldn't this frame marked as 0x00000100 as shown per nft trace ?
Second: I add a table in rt_tables and did
#ip -6 rule add from all table <mytable>
#ip -6 route add default via <mygw> dev wig0
#ip -6 route flush cache
Output
#ip -6 rule sh
0: from all lookup local
32765: from all fwmark 0x100 lookup <mytable>
32766: from all lookup main
but my packet are still going out using enp35s0 as default route. What
am I missing knowing that ct established,related is setted in nftables
rules.
Thanks for any hint
--
Daniel
