netns is a utility that allows you to quickly setup a network namespace. It was written for the purpose of capturing network traffic from a single application (using tcpdump / dumpcap).
When a network namespace is started, all commands used to set this up are printed.
To execute commands inside this network namespace, use the netns exec command
which will use socat to make the external DBus session visible inside this
application (see env.sh). Adjust that file as needed.
Start network namespace:
peter@al:~$ sudo ~/netns/netns 0 start
# ip netns add netns0
# ip link add veth0 type veth peer name veth1
# ip link set veth1 netns netns0
# ip link set veth0 up
# ip addr add 10.9.0.1/24 dev veth0
# ip netns exec netns0 ip link set veth1 up
# ip netns exec netns0 ip addr add 10.9.0.2/24 dev veth1
# ip netns exec netns0 ip route add default via 10.9.0.1 dev veth1
# iptables -t nat -A POSTROUTING -o veth0 -j MASQUERADE
# iptables -A FORWARD -i veth0 -j ACCEPT
# iptables -A FORWARD -o veth0 -j ACCEPT
# Done!
peter@al:~$ ip addr show veth0
14: veth0@if13: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 36:13:bd:c5:2f:e8 brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet 10.9.0.1/24 scope global veth0
valid_lft forever preferred_lft forever
inet6 fe80::3413:bdff:fec5:2fe8/64 scope link
valid_lft forever preferred_lft forever
For access to external (LAN or internet) hosts, see below.
Enter the network namespace. It uses sudo to change the user back to the original user:
peter@al:~$ sudo ~/netns/netns 0 exec
(netns0)peter@al:~$ whoami
peter
(netns0)peter@al:~$ ip addr show veth1
13: veth1@if14: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 26:01:ad:ba:a1:ee brd ff:ff:ff:ff:ff:ff link-netnsid 0
inet 10.9.0.2/24 scope global veth1
valid_lft forever preferred_lft forever
inet6 fe80::2401:adff:feba:a1ee/64 scope link
valid_lft forever preferred_lft forever
Run it without arguments to get usage information:
peter@al:~$ ~/netns/netns
Usage: netns ns-no [dry-]{start|stop}
netns ns-no status
netns ns-no exec [command [command args]]
The namespace number must be between 0 and 255 (inclusive)
For namespace number 4, the layout will be:
(host) veth8 (10.9.4.1)
|
[ netns: ns4 ]
|
(namespace) veth9 (10.9.4.2)
The default setup only facilitates communication between the network namespace and the main host where the commands are run. To enable access from the network namespace to external hosts, you must allow IP forwarding and enable NAT for outgoing packets over the default interface (e.g. eth0, ens3 or wlan0). This has to be done only once (changes are persisted until the next reboot):
sudo sysctl net.ipv4.ip_forward=1
sudo iptables -t nat -A POSTROUTING -o ens3 -j MASQUERADE
On some systems, the default DNS servers point to a local resolver (127.0.0.53
on Ubuntu 18.04). As the network namespace does not have such a local resolver,
you should probably change /etc/resolv.conf to something like:
nameserver 1.1.1.1
To help you identify whether your shell is in a namespace, you can look at the
output of ip link.
For your convenience, you can also make the prompt display the network namespace
name by putting this in your ~/.bashrc:
_ns_name=$(ip netns identify 2>/dev/null)
PS1=${_ns_name:+(${_ns_name})}${PS1}
unset _ns_name
To use the ip netns identify command as a regular user, the permissions of
/var/run/netns need to be adjusted. For example:
sudo setfacl -m u:$USER:rx /var/run/netns
"Namespaces in operation, part 7: Network namespaces [LWN.net]" by Jake Edge, January 22, 2014. Retrieved February 14, 2015. https://lwn.net/Articles/580893/