Ευάγγελος Μπαλάσκας - Network Namespaces - Part One » Evaggelos Balaskas

May

2020

Network Namespaces - Part One

Posted by ebal at 08:47:13 in blog, planet_ellak, planet_Sysadmin, planet_fsfe

Have you ever wondered how containers work on the network level? How they isolate resources and network access? Linux namespaces is the magic behind all these and in this blog post, I will try to explain how to setup your own private, isolated network stack on your linux box.

notes based on ubuntu:20.04, root access.

current setup

Our current setup is similar to this

ip-netns00

List ethernet cards

ip address list

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 52:54:00:ea:50:87 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.80/24 brd 192.168.122.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::5054:ff:feea:5087/64 scope link
       valid_lft forever preferred_lft forever

List routing table

ip route list

default via 192.168.122.1 dev eth0 proto static
192.168.122.0/24 dev eth0 proto kernel scope link src 192.168.122.80

ip-netns01

Checking internet access and dns

ping -c 5 libreops.cc

PING libreops.cc (185.199.111.153) 56(84) bytes of data.
64 bytes from 185.199.111.153 (185.199.111.153): icmp_seq=1 ttl=54 time=121 ms
64 bytes from 185.199.111.153 (185.199.111.153): icmp_seq=2 ttl=54 time=124 ms
64 bytes from 185.199.111.153 (185.199.111.153): icmp_seq=3 ttl=54 time=182 ms
64 bytes from 185.199.111.153 (185.199.111.153): icmp_seq=4 ttl=54 time=162 ms
64 bytes from 185.199.111.153 (185.199.111.153): icmp_seq=5 ttl=54 time=168 ms

--- libreops.cc ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4004ms
rtt min/avg/max/mdev = 121.065/151.405/181.760/24.299 ms

linux network namespace management

In this article we will use the below programs:

so, let us start working with network namespaces.

list

To view the network namespaces, we can type:

ip netns
ip netns list

This will return nothing, an empty list.

help

So quicly view the help of ip-netns

# ip netns help

Usage:  ip netns list
  ip netns add NAME
  ip netns attach NAME PID
  ip netns set NAME NETNSID
  ip [-all] netns delete [NAME]
  ip netns identify [PID]
  ip netns pids NAME
  ip [-all] netns exec [NAME] cmd ...
  ip netns monitor
  ip netns list-id [target-nsid POSITIVE-INT] [nsid POSITIVE-INT]
NETNSID := auto | POSITIVE-INT

monitor

To monitor in real time any changes, we can open a new terminal and type:

ip netns monitor

Add a new namespace

ip netns add ebal

List namespaces

ip netns list

root@ubuntu2004:~# ip netns add ebal
root@ubuntu2004:~# ip netns list
ebal

We have one namespace

Delete Namespace

ip netns del ebal

Full example

root@ubuntu2004:~# ip netns
root@ubuntu2004:~# ip netns list
root@ubuntu2004:~# ip netns add ebal
root@ubuntu2004:~# ip netns list
ebal
root@ubuntu2004:~# ip netns
ebal
root@ubuntu2004:~# ip netns del ebal
root@ubuntu2004:~#
root@ubuntu2004:~# ip netns
root@ubuntu2004:~# ip netns list
root@ubuntu2004:~#

monitor

root@ubuntu2004:~# ip netns monitor
add ebal
delete ebal

exec

We can run commands inside a namespace.

eg.

ip netns exec ebal ip a

root@ubuntu2004:~# ip netns exec ebal ip a
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

bash

we can also open a shell inside the namespace and run commands throught the shell.
eg.

root@ubuntu2004:~# ip netns exec ebal bash

root@ubuntu2004:~# ip a
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

root@ubuntu2004:~# exit
exit

ip-netns02

as you can see, the namespace is isolated from our system. It has only one local interface and nothing else.

We can bring up the loopback interface up

root@ubuntu2004:~# ip link set lo up

root@ubuntu2004:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever

root@ubuntu2004:~# ip r

veth

The veth devices are virtual Ethernet devices. They can act as tunnels between network namespaces to create a bridge to a physical network device in another namespace, but can also be used as standalone network devices.

Think Veth as a physical cable that connects two different computers. Every veth is the end of this cable.

So we need 2 virtual interfaces to connect our system and the new namespace.

ip link add v-eth0 type veth peer name v-ebal

ip-netns03

eg.

root@ubuntu2004:~# ip link add v-eth0 type veth peer name v-ebal

root@ubuntu2004:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 52:54:00:ea:50:87 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.80/24 brd 192.168.122.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::5054:ff:feea:5087/64 scope link
       valid_lft forever preferred_lft forever

3: v-ebal@v-eth0: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether d6:86:88:3f:eb:42 brd ff:ff:ff:ff:ff:ff

4: v-eth0@v-ebal: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether 3e:85:9b:dd:c7:96 brd ff:ff:ff:ff:ff:ff

Attach veth0 to namespace

Now we are going to move one virtual interface (one end of the cable) to the new network namespace

ip link set v-ebal netns ebal

ip-netns03

we will see that the interface is not showing on our system

root@ubuntu2004:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 52:54:00:ea:50:87 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.80/24 brd 192.168.122.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::5054:ff:feea:5087/64 scope link
       valid_lft forever preferred_lft forever

4: v-eth0@if3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether 3e:85:9b:dd:c7:96 brd ff:ff:ff:ff:ff:ff link-netns ebal

inside the namespace

root@ubuntu2004:~# ip netns exec ebal ip a 

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever

3: v-ebal@if4: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether d6:86:88:3f:eb:42 brd ff:ff:ff:ff:ff:ff link-netnsid 0

Connect the two virtual interfaces

outside

ip addr add 10.10.10.10/24 dev v-eth0

root@ubuntu2004:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 52:54:00:ea:50:87 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.80/24 brd 192.168.122.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::5054:ff:feea:5087/64 scope link
       valid_lft forever preferred_lft forever

4: v-eth0@if3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether 3e:85:9b:dd:c7:96 brd ff:ff:ff:ff:ff:ff link-netns ebal
    inet 10.10.10.10/24 scope global v-eth0
       valid_lft forever preferred_lft forever

inside

ip netns exec ebal ip addr add 10.10.10.20/24 dev v-ebal

root@ubuntu2004:~# ip netns exec ebal ip a 

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever

3: v-ebal@if4: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether d6:86:88:3f:eb:42 brd ff:ff:ff:ff:ff:ff link-netnsid 0
    inet 10.10.10.20/24 scope global v-ebal
       valid_lft forever preferred_lft forever

Both Interfaces are down!

But both interfaces are down, now we need to set up both interfaces:

outside

ip link set v-eth0 up

root@ubuntu2004:~# ip link set v-eth0 up 

root@ubuntu2004:~# ip link show v-eth0
4: v-eth0@if3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state LOWERLAYERDOWN mode DEFAULT group default qlen 1000
    link/ether 3e:85:9b:dd:c7:96 brd ff:ff:ff:ff:ff:ff link-netns ebal

inside

ip netns exec ebal ip link set v-ebal up

root@ubuntu2004:~# ip netns exec ebal ip link set v-ebal up

root@ubuntu2004:~# ip netns exec ebal ip link show v-ebal
3: v-ebal@if4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000
    link/ether d6:86:88:3f:eb:42 brd ff:ff:ff:ff:ff:ff link-netnsid 0

did it worked?

Let’s first see our routing table:

outside

root@ubuntu2004:~# ip r
default via 192.168.122.1 dev eth0 proto static
10.10.10.0/24 dev v-eth0 proto kernel scope link src 10.10.10.10
192.168.122.0/24 dev eth0 proto kernel scope link src 192.168.122.80

inside

root@ubuntu2004:~# ip netns exec ebal ip r
10.10.10.0/24 dev v-ebal proto kernel scope link src 10.10.10.20

Ping !

outside

root@ubuntu2004:~# ping -c 5 10.10.10.20
PING 10.10.10.20 (10.10.10.20) 56(84) bytes of data.
64 bytes from 10.10.10.20: icmp_seq=1 ttl=64 time=0.028 ms
64 bytes from 10.10.10.20: icmp_seq=2 ttl=64 time=0.042 ms
64 bytes from 10.10.10.20: icmp_seq=3 ttl=64 time=0.052 ms
64 bytes from 10.10.10.20: icmp_seq=4 ttl=64 time=0.042 ms
64 bytes from 10.10.10.20: icmp_seq=5 ttl=64 time=0.071 ms

--- 10.10.10.20 ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4098ms
rtt min/avg/max/mdev = 0.028/0.047/0.071/0.014 ms

inside

root@ubuntu2004:~# ip netns exec ebal bash
root@ubuntu2004:~#
root@ubuntu2004:~# ping -c 5 10.10.10.10
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.
64 bytes from 10.10.10.10: icmp_seq=1 ttl=64 time=0.046 ms
64 bytes from 10.10.10.10: icmp_seq=2 ttl=64 time=0.042 ms
64 bytes from 10.10.10.10: icmp_seq=3 ttl=64 time=0.041 ms
64 bytes from 10.10.10.10: icmp_seq=4 ttl=64 time=0.044 ms
64 bytes from 10.10.10.10: icmp_seq=5 ttl=64 time=0.053 ms

--- 10.10.10.10 ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4088ms
rtt min/avg/max/mdev = 0.041/0.045/0.053/0.004 ms
root@ubuntu2004:~# exit
exit

It worked !!

ip-netns03

End of part one.

Tag(s): linux, namespaces, network, ip-netns, veth, iproute2

Evaggelos Balaskas - System Engineer

current setup

List ethernet cards

List routing table

Checking internet access and dns

linux network namespace management

list

help

monitor

Add a new namespace

List namespaces

Delete Namespace

Full example

Directory

exec

bash

We can bring up the loopback interface up

veth

Attach veth0 to namespace

inside the namespace

Connect the two virtual interfaces

Both Interfaces are down!

did it worked?

Ping !

Admin area

Categories

Archives

Evaggelos Balaskas - System Engineer

current setup

List ethernet cards

List routing table

Checking internet access and dns

linux network namespace management

list

help

monitor

Add a new namespace

List namespaces

Delete Namespace

Full example

Directory

exec

bash

We can bring up the loopback interface up

veth

Attach veth0 to namespace

inside the namespace

Connect the two virtual interfaces

Both Interfaces are down!

did it worked?

Ping !

Search

Admin area

Categories

Archives