Enable MACVLAN

MacVLAN type definition

Create Network Attach-Definition macvlan type

YAML
    
 
apiVersion: "k8s.cni.cncf.io/v1"kind: NetworkAttachmentDefinitionmetadata:  name: macvlan-conf-1spec:  config: '{      "cniVersion": "0.3.0",      "name": "macvlan-conf-1",      "type": "macvlan",      "master": "eth1",      "mode": "bridge",      "ipam": {        "type": "whereabouts",        "range": "192.168.60.0/24",        "range_start": "192.168.60.20",        "range_end": "192.168.60.50",        "gateway": "192.168.60.1"      }    }'
Copy

The master key value is the reference to the *_second nic *_in our worker nodes.

Bash
    
 
$ kubectl apply -f networkattachdefinition.yaml
Copy

Network Attach Definition Validation macvlan type

Let’s validate our work by listing and describing our new Network Attach Definition

Bash
    
 
$ $ kubectl get net-attach-defNAME               AGEmacvlan-conf-1     134m
Copy

YAML
    
# kubectl describe net-attach-def macvlan-conf-1Name:         macvlan-conf-1Namespace:    defaultLabels:       <none>Annotations:  kubectl.kubernetes.io/last-applied-configuration:                {"apiVersion":"k8s.cni.cncf.io/v1","kind":"NetworkAttachmentDefinition","metadata":{"annotations":{},"name":"macvlan-conf-1","namespace":"...API Version:  k8s.cni.cncf.io/v1Kind:         NetworkAttachmentDefinitionMetadata:  Creation Timestamp:  2020-09-18T13:38:22Z  Generation:          2  Resource Version:    6958535  Self Link:           /apis/k8s.cni.cncf.io/v1/namespaces/default/network-attachment-definitions/macvlan-conf-1  UID:                 99a902e1-8fd1-4f37-81d7-229361029455Spec:  Config:  { "cniVersion": "0.3.0", "type": "macvlan", "master": "eth1", "mode": "bridge", "ipam": { "type": "whereabouts", "range": "192.168.60.0/24", "range_start": "192.168.60.20", "range_end": "192.168.60.50", "gateway": "192.168.60.1", "routes": [{"dst": "0.0.0.0/0"}], "gateway": "192.168.60.1" } }Events:    <none>
Copy

Pod Definitions with 2 interfaces

In order for pods to be created with an additional nic, and the same principle applies for adding more than two NICs to the pod, the pod definition should be called with network annotations making reference to the Network Attach Definition(s) please use the following pods definitions to create a testbed.

Create Pods with mcvlan interfaces

YAML
    
 
apiVersion: v1kind: Podmetadata:  name: pod0-case-01  annotations:    k8s.v1.cni.cncf.io/networks: macvlan-conf-1spec:  containers:  - name: pod0-case-01    image: docker.io/centos/tools:latest    command:    - /sbin/init
Copy

YAML
    
 
apiVersion: v1kind: Podmetadata:  name: pod1-case-01  annotations:    k8s.v1.cni.cncf.io/networks: macvlan-conf-1spec:  containers:  - name: pod1-case-01    image: docker.io/centos/tools:latest    command:    - /sbin/init
Copy

Deploy the new pods

Bash
    
 
$ kubectl apply -f pod1-case1.yaml$ kubectl apply -f pod0-case1.yaml
Copy

Validate Pods Creation with macvlan interfaces

Let’s validate your work by confirming that the pods got created with an additional interface by doing the following commands:

Bash
    
$ kubectl get pods -o wideNAME                                      READY   STATUS    RESTARTS   AGE    IP             NODE            NOMINATED NODE   READINESS GATESpod0-case-01                              1/1     Running   0          142m   10.135.41.90   192.168.50.16   <none>           <none>pod1-case-01                              1/1     Running   0          140m   10.135.41.91   192.168.50.16   <none>           <none>
Copy

Bash
    
​x
    
$ kubectl exec -it pod0-case-01 -- ip -d address1: 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 promiscuity 0 numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535    inet 127.0.0.1/8 scope host lo       valid_lft forever preferred_lft forever3: eth0@if585: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1472 qdisc noqueue state UP group default    link/ether 16:66:b2:53:e4:f8 brd ff:ff:ff:ff:ff:ff link-netnsid 0 promiscuity 0    veth numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535    inet 10.135.41.90/24 brd 10.135.41.255 scope global eth0       valid_lft forever preferred_lft forever4: net1@if401: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default    link/ether 02:5b:b8:96:0a:d7 brd ff:ff:ff:ff:ff:ff link-netnsid 0 promiscuity 0    macvlan mode bridge numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535    inet 192.168.60.20/24 brd 192.168.60.255 scope global net1       valid_lft forever preferred_lft forever​$ kubectl exec -it pod1-case-01 -- ip -d address1: 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 promiscuity 0 numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535    inet 127.0.0.1/8 scope host lo       valid_lft forever preferred_lft forever3: eth0@if586: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1472 qdisc noqueue state UP group default    link/ether fa:61:88:fb:53:da brd ff:ff:ff:ff:ff:ff link-netnsid 0 promiscuity 0    veth numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535    inet 10.135.41.91/24 brd 10.135.41.255 scope global eth0       valid_lft forever preferred_lft forever4: net1@if401: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default    link/ether ca:c3:73:0e:f3:e5 brd ff:ff:ff:ff:ff:ff link-netnsid 0 promiscuity 0    macvlan mode bridge numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535    inet 192.168.60.21/24 brd 192.168.60.255 scope global net1       valid_lft forever preferred_lft forever
Copy

Connectivity Testing

East-West macvlan traffic

Bash
    
 
$ kubectl exec -it pod0-case-01 -- ping -c 5 192.168.60.21PING 192.168.60.21 (192.168.60.21) 56(84) bytes of data.64 bytes from 192.168.60.21: icmp_seq=1 ttl=64 time=0.289 ms64 bytes from 192.168.60.21: icmp_seq=2 ttl=64 time=0.061 ms64 bytes from 192.168.60.21: icmp_seq=3 ttl=64 time=0.059 ms64 bytes from 192.168.60.21: icmp_seq=4 ttl=64 time=0.062 ms64 bytes from 192.168.60.21: icmp_seq=5 ttl=64 time=0.058 ms​--- 192.168.60.21 ping statistics ---5 packets transmitted, 5 received, 0% packet loss, time 4001msrtt min/avg/max/mdev = 0.058/0.105/0.289/0.092 ms
Copy

Bash
    
 
$ kubectl exec -it pod1-case-01 -- ping -c 5 192.168.60.20PING 192.168.60.20 (192.168.60.20) 56(84) bytes of data.64 bytes from 192.168.60.20: icmp_seq=1 ttl=64 time=0.253 ms64 bytes from 192.168.60.20: icmp_seq=2 ttl=64 time=0.044 ms64 bytes from 192.168.60.20: icmp_seq=3 ttl=64 time=0.050 ms64 bytes from 192.168.60.20: icmp_seq=4 ttl=64 time=0.047 ms64 bytes from 192.168.60.20: icmp_seq=5 ttl=64 time=0.069 ms​--- 192.168.60.20 ping statistics ---5 packets transmitted, 5 received, 0% packet loss, time 4000msrtt min/avg/max/mdev = 0.044/0.092/0.253/0.081 ms
Copy

North-South macvlan traffic:

Bash
    
 
$ kubectl exec -it pod0-case-01 -- ping -c 5 192.168.60.1PING 192.168.60.1 (192.168.60.1) 56(84) bytes of data.64 bytes from 192.168.60.1: icmp_seq=1 ttl=64 time=0.925 ms64 bytes from 192.168.60.1: icmp_seq=2 ttl=64 time=0.536 ms64 bytes from 192.168.60.1: icmp_seq=3 ttl=64 time=0.483 ms64 bytes from 192.168.60.1: icmp_seq=4 ttl=64 time=0.640 ms64 bytes from 192.168.60.1: icmp_seq=5 ttl=64 time=0.488 ms
Copy

Bash
    
 
$ kubectl exec -it pod1-case-01 -- ping -c 5 192.168.60.1PING 192.168.60.1 (192.168.60.1) 56(84) bytes of data.64 bytes from 192.168.60.1: icmp_seq=1 ttl=64 time=1.31 ms64 bytes from 192.168.60.1: icmp_seq=2 ttl=64 time=0.567 ms64 bytes from 192.168.60.1: icmp_seq=3 ttl=64 time=0.473 ms64 bytes from 192.168.60.1: icmp_seq=4 ttl=64 time=0.605 ms64 bytes from 192.168.60.1: icmp_seq=5 ttl=64 time=0.647 ms​--- 192.168.60.1 ping statistics ---5 packets transmitted, 5 received, 0% packet loss, time 4002msrtt min/avg/max/mdev = 0.473/0.721/1.313/0.301 ms
Copy

Last updated on Jan 23, 2026

Was this page helpful?