Macvtap binding

Overview

With the macvtap binding plugin, virtual machines are directly exposed to the Kubernetes nodes L2 network. This is achieved by 'extending' an existing network interface with a virtual device that has its own MAC address.

Its main benefits are:

• Direct connection to the node nic with no intermediate bridges.

Functionality support

Functionality	Support
Run without extra capabilities (on pod)	Yes
Migration support	No
IPAM support (on pod)	No
Primary network (pod network)	No
Secondary network	Yes

Known Issues

• Live migration is not fully supported, see issue #5912

Warning: On KinD clusters, the user needs to adjust the cluster configuration, mounting dev of the running host onto the KinD nodes, because of a known issue.

Deployment

The macvtap solution consists of a CNI and a DP.

In order to use macvtap, the following points need to be covered:

• Deploy the CNI plugin binary on the nodes.
• Deploy the Device Plugin daemon on the nodes.
• Configure which node interfaces are exposed.
• Define a NetworkAttachmentDefinition that points to the CNI plugin.

Macvtap CNI and DP deployment on nodes

To simplify the procedure, use the Cluster Network Addons Operator to deploy and configure the macvtap components in your cluster.

The aforementioned operator effectively deploys the macvtap cni and device plugin.

Expose node interface to the macvtap device plugin

There are two different alternatives to configure which host interfaces get exposed to the user, enabling them to create macvtap interfaces on top of:

• select the host interfaces: indicates which host interfaces are exposed.
• expose all interfaces: all interfaces of all hosts are exposed.

Both options are configured via the macvtap-deviceplugin-config ConfigMap, and more information on how to configure it can be found in the macvtap-cni repo.

This is a minimal example, in which the eth0 interface of the Kubernetes nodes is exposed, via the lowerDevice attribute.

kind: ConfigMap
apiVersion: v1
metadata:
  name: macvtap-deviceplugin-config
data:
  DP_MACVTAP_CONF: |
    [
        {
            "name"        : "dataplane",
            "lowerDevice" : "eth0",
            "mode"        : "bridge",
            "capacity"    : 50
        },
    ]

This step can be omitted, since the default configuration of the aforementioned ConfigMap is to expose all host interfaces (which is represented by the following configuration):

kind: ConfigMap
apiVersion: v1
metadata:
  name: macvtap-deviceplugin-config
data:
  DP_MACVTAP_CONF: '[]'

Macvtap NetworkAttachmentDefinition

The configuration needed for a macvtap network attachment can be minimalistic:

kind: NetworkAttachmentDefinition
apiVersion: k8s.cni.cncf.io/v1
metadata:
  name: macvtapnetwork
  annotations:
    k8s.v1.cni.cncf.io/resourceName: macvtap.network.kubevirt.io/eth0
spec:
  config: '{
      "cniVersion": "0.3.1",
      "name": "macvtapnetwork",
      "type": "macvtap",
      "mtu": 1500
    }'

The object should be created in a "default" namespace where all other namespaces can access, or, in the same namespace the VMs reside in.

The requested k8s.v1.cni.cncf.io/resourceName annotation must point to an exposed host interface (via the lowerDevice attribute, on the macvtap-deviceplugin-config ConfigMap).

Macvtap network binding plugin

[v1.1.1]

The binding plugin replaces the experimental core macvtap binding implementation (including its API).

Note: The network binding plugin infrastructure and the macvtap plugin specifically are in Alpha stage. Please use them with care, preferably on a non-production deployment.

The macvtap binding plugin consists of the following components:

• Macvtap CNI plugin.

The plugin needs to:

• Enable the network binding plugin framework FG.
• Register the binding plugin on the Kubevirt CR.
• Reference the network binding by name from the VM spec interface.

And in detail:

Feature Gate

If not already set, add the NetworkBindingPlugins FG.

kubectl patch kubevirts -n kubevirt kubevirt --type=json -p='[{
  "op": "add",
  "path": "/spec/configuration/developerConfiguration/featureGates/-",
  "value": "NetworkBindingPlugins"
}]'

Note: The specific macvtap plugin has no FG by its own. It is up to the cluster admin to decide if the plugin is to be available in the cluster. The macvtap binding is still in evaluation, use it with care.

Macvtap Registration

The macvtap binding plugin configuration needs to be added to the kubevirt CR in order to be used by VMs.

To register the macvtap binding, patch the kubevirt CR as follows:

kubectl patch kubevirts -n kubevirt kubevirt --type=json -p='[{"op": "add", "path": "/spec/configuration/network",   "value": {
            "binding": {
                "macvtap": {
                    "domainAttachmentType": "tap"
                }
            }
        }}]'

VM Macvtap Network Interface

Set the VM network interface binding name to reference the one defined in the kubevirt CR.

---
apiVersion: kubevirt.io/v1
kind: VirtualMachine
metadata:
  labels:
    kubevirt.io/vm: vm-net-binding-macvtap
  name: vm-net-binding-macvtap
spec:
  runStrategy: Always
  template:
    metadata:
      labels:
        kubevirt.io/vm: vm-net-binding-macvtap
    spec:
      domain:
        devices:
          disks:
          - disk:
              bus: virtio
            name: containerdisk
          - disk:
              bus: virtio
            name: cloudinitdisk
          interfaces:
          - name: podnet
            masquerade: {}
          - name: hostnetwork
            binding:
              name: macvtap
          rng: {}
        resources:
          requests:
            memory: 1024M
      networks:
      - name: podnet
        pod: {}
      - name: hostnetwork
        multus:
          networkName: macvtapnetwork
      terminationGracePeriodSeconds: 0
      volumes:
      - containerDisk:
          image: quay.io/kubevirt/fedora-with-test-tooling-container-disk:v1.1.0
        name: containerdisk
      - cloudInitNoCloud:
          networkData: |
            version: 2
            ethernets:
              eth0:
                dhcp4: true
        name: cloudinitdisk

The multus networkName value should correspond with the name used in the network attachment definition section.

The binding value should correspond with the name used in the registration.