You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I switched to Talos from k3s, and upon trying to use pod-gateway, it instead crashes.
What did you expect to happen:
I did not expect it to choke on the K8S_DNS part when that resolves to v4 and v6.
Anything else you would like to add:
Talos version 1.7.5
Kubernetes: 1.30.2
Additional Information:
Defaulted container "vpn-gateway-pod-gateway" out of: vpn-gateway-pod-gateway, gluetun, netshoot, routes (init)
+ cat /default_config/settings.sh
#!/bin/bash
# hostname of the gateway - it must accept vxlan and DHCP traffic
# clients get it as env variable
GATEWAY_NAME="$gateway"
# K8S DNS IP address
# clients get it as env variable
K8S_DNS_IPS="$K8S_DNS_ips"
# Blank sepated IPs not sent to the POD gateway but to the default K8S
# This is needed, for example, in case your CNI does
# not add a non-default rule for the K8S addresses (Flannel does)
NOT_ROUTED_TO_GATEWAY_CIDRS=""
# Vxlan ID to use
VXLAN_ID="42"
# VXLAN need an /24 IP range not conflicting with K8S and local IP ranges
VXLAN_IP_NETWORK="172.16.0"
# Keep a range of IPs for static assignment in nat.conf
VXLAN_GATEWAY_FIRST_DYNAMIC_IP=20
# If using a VPN, interface name created by it
VPN_INTERFACE=tun0
# Prevent non VPN traffic to leave the gateway
VPN_BLOCK_OTHER_TRAFFIC=true
# If VPN_BLOCK_OTHER_TRAFFIC is true, allow VPN traffic over this port
VPN_TRAFFIC_PORT=443
# Traffic to these IPs will be send through the K8S gateway
VPN_LOCAL_CIDRS="10.0.0.0/8 192.168.0.0/16"
# DNS queries to these domains will be resolved by K8S DNS instead of
# the default (typcally the VPN client changes it)
DNS_LOCAL_CIDRS="local"
# dnsmasq monitors directories. /etc/resolv.conf in a container is in another
# file system so it does not work. To circumvent this a copy is made using
# inotifyd
RESOLV_CONF_COPY=/etc/resolv_copy.conf
# ICMP heartbeats are used to ensure the pod-gateway is connectable from the clients.
# The following value can be used to to provide more stability in an unreliable network connection.
CONNECTION_RETRY_COUNT=1
# you want to disable DNSSEC with the gateway then set this to false
GATEWAY_ENABLE_DNSSEC=true
# If you use nftables for iptables you need to set this to yes
IPTABLES_NFT=no
# Set to WAN/VPN IP to enable SNAT instead of Masquerading
SNAT_IP=""
# Set the VPN MTU. It also adjust the VXLAN MTU to avoid fragmenting the package in the gateway (VXLAN-> MTU)
VPN_INTERFACE_MTU=""
+ . /default_config/settings.sh
++ GATEWAY_NAME=
++ K8S_DNS_IPS=
++ NOT_ROUTED_TO_GATEWAY_CIDRS=
++ VXLAN_ID=42
++ VXLAN_IP_NETWORK=172.16.0
++ VXLAN_GATEWAY_FIRST_DYNAMIC_IP=20
++ VPN_INTERFACE=tun0
++ VPN_BLOCK_OTHER_TRAFFIC=true
++ VPN_TRAFFIC_PORT=443
++ VPN_LOCAL_CIDRS='10.0.0.0/8 192.168.0.0/16'
++ DNS_LOCAL_CIDRS=local
++ RESOLV_CONF_COPY=/etc/resolv_copy.conf
++ CONNECTION_RETRY_COUNT=1
++ GATEWAY_ENABLE_DNSSEC=true
++ IPTABLES_NFT=no
++ SNAT_IP=
++ VPN_INTERFACE_MTU=
+ cat /config/settings.sh
#!/bin/sh
# Generated by vpn-gateway-pod-gateway
DNS_LOCAL_CIDRS="local"
NOT_ROUTED_TO_GATEWAY_CIDRS="10.0.0.0/8 192.168.0.0/16"
VPN_BLOCK_OTHER_TRAFFIC="false"
VPN_INTERFACE="wg0"
VPN_LOCAL_CIDRS="10.0.0.0/8 192.168.0.0/16"
VPN_TRAFFIC_PORT="51820"
VXLAN_GATEWAY_FIRST_DYNAMIC_IP="20"
VXLAN_ID="44"
VXLAN_IP_NETWORK="172.16.1"
+ . /config/settings.sh
++ DNS_LOCAL_CIDRS=local
++ NOT_ROUTED_TO_GATEWAY_CIDRS='10.0.0.0/8 192.168.0.0/16'
++ VPN_BLOCK_OTHER_TRAFFIC=false
++ VPN_INTERFACE=wg0
++ VPN_LOCAL_CIDRS='10.0.0.0/8 192.168.0.0/16'
++ VPN_TRAFFIC_PORT=51820
++ VXLAN_GATEWAY_FIRST_DYNAMIC_IP=20
++ VXLAN_ID=44
++ VXLAN_IP_NETWORK=172.16.1
+ '[' '!' -f /etc/resolv.conf.org ']'
+ cp /etc/resolv.conf /etc/resolv.conf.org
/etc/resolv.conf.org written
+ echo '/etc/resolv.conf.org written'
++ grep nameserver /etc/resolv.conf.org
++ cut '-d ' -f2
+ K8S_DNS='10.96.0.10
2001:14ba:74ae:3405::1e:a'
+ cat
+ [[ true == true ]]
+ cat
+ for local_cidr in $DNS_LOCAL_CIDRS
+ cat
+ /bin/copy_resolv.sh
copying /etc/resolv.conf to /etc/resolv_copy.conf
+ dnsmasq=18
+ dnsmasq -k
+ inotifyd=19
+ inotifyd /bin/copy_resolv.sh /etc/resolv.conf:ce
+ trap _kill_procs SIGTERM
+ wait -n
dnsmasq: bad option at line 30 of /etc/dnsmasq.d//pod-gateway.conf
The text was updated successfully, but these errors were encountered:
Details
What steps did you take and what happened:
I switched to Talos from k3s, and upon trying to use pod-gateway, it instead crashes.
What did you expect to happen:
I did not expect it to choke on the K8S_DNS part when that resolves to v4 and v6.
Anything else you would like to add:
Talos version 1.7.5
Kubernetes: 1.30.2
Additional Information:
The text was updated successfully, but these errors were encountered: