Simple utility to simplify creation of network namespaces, interfaces and bridges. Purpose is to create "virtualized" network environment for testing purposes.

Required topology is described in a simple json file. Following example creates 3 network namespaces:

• ns1 - with network interface eth0 (which is veth with other end inside namespace nsb)
• ns2 - with network interface eth0 (which is veth with other end inside namespace nsb)
• nsb - with network interface b0 (which is bridge connecting veth interfaces from ns1 and ns2)

{
  "namespaces": [
    {"name": "ns1"},
    {"name": "ns2"},
    {"name": "nsb"}
  ],

  "interfaces": [
    {"namespace": "ns1", "name": "eth0", "type": "veth", "peer_namespace": "nsb", "peer_name": "c1"},
    {"namespace": "ns2", "name": "eth0", "type": "veth", "peer_namespace": "nsb", "peer_name": "c2"},
    {"namespace": "nsb", "name": "b0", "type": "bridge", "slave": ["c1", "c2"]}
  ],
 
  "ips": [
    {"namespace": "ns1", "interface": "eth0", "address": "1.1.1.1/24"},
    {"namespace": "ns1", "interface": "eth0", "address": "fd00::100/64"},
    {"namespace": "ns2", "interface": "eth0", "address": "1.1.1.2/24"},
    {"namespace": "ns2", "interface": "eth0", "address": "fd00::101/64"}
  ],
  "exec": [
    {"namespace": "ns1", "command": "ip route add 10.0.1.0/24 via 10.0.0.20"}
  ]
}

+-------   +--------+   +------+
|  ns1 |   |   nsb  |   |  ns2 |
|      |   |        |   |      |
|  eth0-----c1-b0-c2-----eth0  |
+------+   +--------+   +------+

To compile utility just clone repository and run go build. Or run just go build github.com/tom-code/nsutil

Store example configuration to config.json and run: ./nsutil create. You can then check what is inside network namespaces and try to ping other ends:

ip netns exec ns1 ip addr
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: eth0@if2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default 
    link/ether 2a:25:6b:c9:51:86 brd ff:ff:ff:ff:ff:ff link-netns nsb
    inet 1.1.1.1/24 brd 1.1.1.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fd00::100/64 scope global 
       valid_lft forever preferred_lft forever
    inet6 fe80::2825:6bff:fec9:5186/64 scope link 
       valid_lft forever preferred_lft forever

ip netns exec ns1 ping fd00::101
PING fd00::101(fd00::101) 56 data bytes
64 bytes from fd00::101: icmp_seq=1 ttl=64 time=0.115 ms

ip netns exec ns1 ping 1.1.1.2  
PING 1.1.1.2 (1.1.1.2) 56(84) bytes of data.
64 bytes from 1.1.1.2: icmp_seq=2 ttl=64 time=0.069 ms

To cleanup resources created by nsutil run: nsutil delete. Utility will attempt to delete all resources it did create.

• In basic form new named namespace is created: {"name": "ns1"}
• It is also possible to use existing namespace of existing container using this syntax: {"name": "ns3", "type": "container", "container-id": "3c521aadf445"}
• To use existing namespace attached to known pid use: {"name": "ns4", "type": "pid", "pid": "4144225"}

supported interfaces are:

To create veth interface use: {"namespace": "ns1", "name": "eth9", "type": "veth", "peer_namespace": "nsb", "peer_name": "c1"}. This example will create veth pair with one end in namespace "ns1" and interface named "eth2". Other end will be in namespace "nsb" with interface named "c1".

To create bridge use for example {"namespace": "nsb", "name": "b0", "type": "bridge", "slave": ["c1", "c2", "c3"]}. This example will create bridge named b0 in namespace nsb. Interfaces c1, c2, c3 will be attached to this bridge.

To create macvlan interface use for example {"namespace": "ns1", "name": "eth0", "type": "macvlan", "parent": "eth0"}. This example will create macvlan interface named eth0 in namespace ns1. Parent interface is eth0 from root namespace.

To create ipvlan interface use for example {"namespace": "ns1", "name": "eth0", "type": "ipvlan", "parent": "eth0"}. This example will create ipvlan interface named eth0 in namespace ns1. Parent interface is eth0 from root namespace.

To create vlan interface based on existing interface use for example {"namespace": "ns1", "name": "eth0", "type": "vlan", "vlan_id": 100}. This example will create vlan interface named eth0-100 in namespace ns1 based don eth0.