This guide provides step-by-step instructions for setting up \texttt{dnsmasq} as a DNS and DHCP server in an Incus container running Debian. The network configuration is managed using Netplan and a custom Python script to create virtual Ethernet (veth) pairs and bridges, ensuring proper network integration.
The \texttt{security.syscalls.intercept.mount}, \texttt{security.nesting}, and \texttt{security.privileged} settings are required for \texttt{dnsmasq} and Docker to function correctly in the container.
\subsection{Installing Additional Packages}
Install the necessary packages inside the container:
\begin{lstlisting}[language=bash]
incus exec dnsmasq-container -- apt update
incus exec dnsmasq-container -- apt install -y \
netplan.io \
sudo vim nano git tmux mc zip unzip curl wget htop lynx \
\subsection{Configuring Network with Veth Pairs and Netplan}
To enable advanced networking, use the provided Python script (\texttt{link.py}) to create a virtual Ethernet (veth) pair connecting the container to the host's network namespace, with an optional bridge for network integration. Save the following script as \texttt{link.py} on the host:
\begin{lstlisting}[language=python]
import argparse
import os
import subprocess
import sys
from pyroute2 import IPRoute, NetNS
# ... (rest of the link.py script as provided) ...
\end{lstlisting}
Run the script to create a veth pair, move one end to the container's network namespace, and attach it to a bridge on the host. First, identify the container's name or ID:
\begin{lstlisting}[language=bash]
incus list
\end{lstlisting}
Assuming the container name is \texttt{dnsmasq-container}, execute the script with sudo privileges:
\item\texttt{--namespace1 dnsmasq-container}: Specifies the container's network namespace (Incus container).
\item\texttt{--namespace2 1}: Specifies the default (host) namespace.
\item\texttt{--name1 veth-container}: Names the veth interface inside the container.
\item\texttt{--name2 veth-host}: Names the veth interface on the host.
\item\texttt{--bridge2 br0}: Attaches the host's veth interface to a bridge named \texttt{br0}.
\item\texttt{--type1 incus}: Indicates that \texttt{namespace1} is an Incus container.
\end{itemize}
Before running the script, ensure the bridge \texttt{br0} exists on the host. Create it if necessary:
\begin{lstlisting}[language=bash]
sudo ip link add name br0 type bridge
sudo ip link set br0 up
\end{lstlisting}
The script exposes the container's network namespace, creates the veth pair, moves \texttt{veth-container} to the container's namespace, attaches \texttt{veth-host} to \texttt{br0}, and brings both interfaces up.
Configure the container's network using Netplan to assign a static IP address to the \texttt{veth-container} interface (aliased as \texttt{eth0} for simplicity). Create or edit the Netplan configuration file at \texttt{/etc/netplan/01-netcfg.yaml} inside the container:
To test DHCP, connect a client device to the same network (via the \texttt{br0} bridge) and verify that it receives an IP address in the range \texttt{192.168.1.100--192.168.1.200}.
This guide configures \texttt{dnsmasq} as a DNS and DHCP server in an Incus container on Debian. The \texttt{link.py} script and Netplan configuration ensure a robust network setup with veth pairs and static IP addressing. For advanced configurations, refer to the \texttt{dnsmasq} documentation (\texttt{man dnsmasq}) and \texttt{pyroute2} documentation.
