SENSE/Multus L2 Path Provisioning

Follow these steps to provision an L2 path using SENSE and Multus.

Step 1: Create the SENSE path:

Create a new instance from the ”+” button in the bottom right corner.

Create New Instance

Select the service profile you want. (These will be provided and already populated by the ESnet SENSE Team).

Select Service Profile

Fill in the fields specifying the configuration for the L2 path you want. (You can view and search for URIs in the `Visualization` tab).

Fill Configuration

uri: The node’s Uniform Resource Identifier. Can be looked up in search. Will contain the NRP node name and corresponding NIC interface in the name. Example: urn:ogf:network:nrp-nautilus.io:2020:node-2-7.sdsc.optiputer.net:enp65s0f1np1.

assign_ip: Whether or not the orchestrator should assign IPs.

vlan_tag: Can be any VLAN that doesn’t already have a path on it. If the user specifies any, the orchestrator will pick an available VLAN tag.

bandwidth.capacity: QoS bandwidth and not an upper cap.

Specify an alias for the instance (name).

Specify Alias

You can view the status (failure/success) and logs after hitting submit.

View Status

Step 1A (Extra): Using the API

You can also acheieve the same results using the sense-o-api, or the sense-o-py-client.

A detailed step-by-step guide on using the Python client can be found here.

Step 2: Configure the Node

You will see the interface(s) show up on the node(s) specified in the URI(s). Example:

nautilus@node-2-7:~$ ifconfig | grep -i vlan
vlan.3110: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 9000
vlan.3110-ifb: flags=195<UP,BROADCAST,RUNNING,NOARP>  mtu 1500

No further configuration on the node is necessary.

Step 3: Create a Multus NetworkAttachmentDefinition

Define a Multus NetworkAttachmentDefinition using the MACVLAN interface. Save the following YAML as network.yaml:

apiVersion: "k8s.cni.cncf.io/v1"
kind: NetworkAttachmentDefinition
metadata:
  name: macvlan-network
  namespace: <namespace>
spec:
  config: '{
    "cniVersion": "0.3.1",
    "type": "macvlan",
    "master": "vlan.3110",
    "mode": "bridge",
    "ipam": {
      "type": "host-local",
      "subnet": "192.168.1.0/24",
      "rangeStart": "192.168.1.1",
      "rangeEnd": "192.168.1.50"
    }
  }'

Apply it:

kubectl apply -f network.yaml

Step 4: Create a Pod

Deploy a pod using the network. Save the following YAML as pod.yaml:

apiVersion: v1
kind: Pod
metadata:
  name: macvlan-pod
  namespace: <namespace>
  annotations:
    k8s.v1.cni.cncf.io/networks: macvlan-network
spec:
  containers:
  - name: test-container
    image: alpine
    command: ["sh", "-c", "sleep 3600"]
    resources:
      limits:
        memory: "256Mi"
        cpu: "500m"
      requests:
        memory: "128Mi"
        cpu: "250m"