/kube5gnfvo

SEFCOM's fork of kube5gnfvo (used by free5gmano/free5gmano) to fix any issues with the deployment part

Primary LanguagePythonApache License 2.0Apache-2.0

Kube5GNfvo

Table of Contents

Kubernetes

Please refer to Bootstrapping clusters with kubeadm to install Kubernetes.

Using Kubernetes version: v1.15.3

Support Kubernetes version: v1.15 to v1.18

on ubuntu
apt-get install -y kubelet=1.15.3-00 kubeadm=1.15.3-00 kubectl=1.15.3-00 --allow-downgrades

Multus

Please refer to Multus Quickstart Installation Guide to install Multus.

Or follow the instructions
cd kube5gnfvo/example/
kubectl apply -f multus-daemonset-pre-1.16.yml

If using v1.16 or higher

cd kube5gnfvo/example/
kubectl apply -f multus-daemonset.yml

OpenvSwitch

Please refer to Open vSwitch Installation Guide to install OpenvSwitch.

Or follow the instructions
apt install openvswitch-switch -y
ovs-vsctl add-br br1

OVS-CNI

Please refer to Kubevirt OVS CNI to install ovs-cni.

Or follow the instructions
cd kube5gnfvo/example/
kubectl apply -f ovs-cni-pre-1.16.yaml

If using v1.16 or higher

cd kube5gnfvo/example/
kubectl apply -f ovs-cni.yaml
Create a NetworkAttachmentDefinition
cat <<EOF >./ovs-net-crd.yaml
apiVersion: "k8s.cni.cncf.io/v1"
kind: NetworkAttachmentDefinition
metadata:
  name: ovs-net
  annotations:
    k8s.v1.cni.cncf.io/resourceName: ovs-cni.network.kubevirt.io/br1
spec:
  config: '{
      "cniVersion": "0.3.1",
      "type": "ovs",
      "bridge": "br1"
    }'
EOF
kubectl apply -f ovs-net-crd.yaml

Etcd Operator

Please refer to Coreos Etcd Operator Readme to install etcd cluster.

Or follow the instructions
cd kube5gnfvo/example/etcd-cluster/rbac/
./create_role.sh
cd ../etcd-pre-1
kubectl apply -f deployment.yaml
(Please make sure that etcdclusters.etcd.database.coreos.com CRD in Kubernetes has been created)
kubectl apply -f ./

If using v1.16 or higher

cd kube5gnfvo/example/etcd-cluster/rbac/
./create_role.sh
cd ..
kubectl apply -f deployment.yaml
(Please make sure that etcdclusters.etcd.database.coreos.com CRD in Kubernetes has been created)
kubectl apply -f ./

Metrics Server

Please refer to Kubernetes Metrics Server Readme to deploy metrics server.

Or follow the instructions
cd kube5gnfvo/example/metrics-server/
kubectl apply -f ./

Node Exporter

Please refer to Coreos Node Exporter Daemonset to deploy node exporter.

Or follow the instructions
cd kube5gnfvo/example/
kubectl apply -f prom-node-exporter.yaml

KubeVirt

Please refer to KubeVirt Quickstart with Minikube to deploy KubeVirt.

Or follow the instructions
cd kube5gnfvo/example/kubevirt/
kubectl apply -f kubevirt-operator.yaml
(Please make sure that virt-operator pod in Kubernetes has been Running)
kubectl apply -f kubevirt-cr.yaml

kubevirt-py

Please refer to KubeVirt Quickstart with Minikube to deploy kubevirt-py.

Or follow the instructions
pip3 install git+https://github.com/yanyan8566/client-python

Quick Start

This section explains an exmaple deployment of Kube5GNfvo in Kubernetes. Required YAML files can be found in directory.

Create a Configmap that is based on a Config of kubernetes cluster

copy the content into a pure text editor such as Notepad firs

cat ~/.kube/config

And copy it from there into kube5gnfvo-configmap.yaml

cat <<EOF >./kube5gnfvo-configmap.yaml

apiVersion: v1
kind: ConfigMap
metadata:
  name: kube5gnfvo-config
data:
  config: |
    ...
      'Here!'
    ...

EOF

kubectl apply -f kube5gnfvo-configmap.yaml

Create kube5gnfvo ServiceAccount

cat <<EOF >./kube5gnfvo-sa.yaml
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1beta1
metadata:
  name: kube5gnfvo
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin
subjects:
- kind: ServiceAccount
  name: kube5gnfvo
  namespace: default
---
apiVersion: v1
kind: ServiceAccount
metadata:
  name: kube5gnfvo
EOF

kubectl apply -f kube5gnfvo-sa.yaml

Deploy Mysql Database

cat <<EOF >./kube5gnfvo-mysql.yaml
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: kube5gnfvo-mysql
spec:
  strategy:
    type: Recreate
  selector:
    matchLabels:
      app: kube5gnfvo-mysql
  template:
    metadata:
      labels:
        app: kube5gnfvo-mysql
    spec:
      containers:
      - image: mysql:5.6
        name: kube5gnfvo-mysql
        env:
        - name: MYSQL_ROOT_PASSWORD
          value: password
        ports:
        - containerPort: 3306
          name: mysql
        volumeMounts:
        - name: kube5gnfvo-mysql
          mountPath: /var/lib/mysql
        volumeMounts:
        - name: mysql-initdb
          mountPath: /docker-entrypoint-initdb.d
      volumes:
      - name: kube5gnfvo-mysql
        persistentVolumeClaim:
          claimName: kube5gnfvo-mysql
      volumes:
      - name: mysql-initdb
        configMap:
          name: mysql-initdb-config
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: kube5gnfvo-mysql
  labels:
    name: kube5gnfvo-mysql
spec:
  capacity:
    storage: 20Gi
  accessModes:
    - ReadWriteOnce
  hostPath:
    type: DirectoryOrCreate
    path: /mnt/kube5gnfvo-mysql
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: kube5gnfvo-mysql
  namespace: default
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 20Gi
  selector:
    matchExpressions:
    - key: name
      operator: In
      values: ["kube5gnfvo-mysql"]
---
apiVersion: v1
kind: Service
metadata:
  name: kube5gnfvo-mysql
spec:
  ports:
  - port: 3306
  selector:
    app: kube5gnfvo-mysql
---
apiVersion: v1
kind: ConfigMap
metadata:
  name: mysql-initdb-config
data:
  initdb.sql: |
    CREATE DATABASE kube5gnfvo;
EOF

kubectl apply -f kube5gnfvo-mysql.yaml

Deploy kube5gnfvo

cat <<EOF >./kube5gnfvo.yaml
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: kube5gnfvo
spec:
  replicas: 1
  selector:
    matchLabels:
      app: kube5gnfvo
  template:
    metadata:
      labels:
        app: kube5gnfvo
    spec:
      serviceAccountName: kube5gnfvo
      containers:
      - image: free5gmano/kube5gnfvo-stage2
        name: kube5gnfvo
        env:
        - name: DATABASE_PASSWORD
          value: "password"
        - name: DATABASE_HOST
          value: "kube5gnfvo-mysql"
        - name: DATABASE_PORT
          value: "3306"
        command: ["/bin/sh","-c"]
        args: ['python3 manage.py migrate && python3 manage.py runserver 0:8000']
        ports:
        - containerPort: 8000
          name: kube5gnfvo
        volumeMounts:
        - name: kube5gnfvo-vnf-package
          mountPath: /root/NSD
          subPath: NSD
        - name: kube5gnfvo-vnf-package
          mountPath: /root/VnfPackage
          subPath: VnfPackage
        - name: kube-config
          mountPath: /root/config
          subPath: config
      volumes:
      - name: kube5gnfvo-vnf-package
        persistentVolumeClaim:
          claimName: kube5gnfvo-pvc
      - name: kube-config
        configMap:
          name: kube5gnfvo-config
          items:
          - key: config
            path: config
---
apiVersion: v1
kind: Service
metadata:
  name: kube5gnfvo
spec:
  type: NodePort
  ports:
  - port: 8000
    nodePort: 30888
  selector:
    app: kube5gnfvo
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: kube5gnfvo-pvc
  namespace: default
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 20Gi
  selector:
    matchExpressions:
    - key: name
      operator: In
      values: ["kube5gnfvo"]
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: kube5gnfvo-pv
  labels:
    name: kube5gnfvo
spec:
  capacity:
    storage: 20Gi
  accessModes:
    - ReadWriteOnce
  hostPath:
    type: DirectoryOrCreate
    path: /mnt/kube5gnfvo
EOF

kubectl apply -f kube5gnfvo.yaml

Example deployments

Testing kube5gnfvo

POSTMAN

We use POSTMAN to send request to API server. You can download at here.

Create NS instance

Compress files

Under kube5gnfvo/example/free5gcv1

You will see two directories, ns and Vnfpackage. Zip ns directory into ns.zip. In kube5gnfvo/example/free5gcv1/Vnfpackage, you will see seven directories, zip each of them into .zip file directly. Now the file structure shown as below:

kube5gnfvo/example/free5gcv1
|--ns
|--ns.zip
|--vnfpackage
   |--amf
   |--hss
   |--mongodb
   |--pcrf
   |--smf
   |--upf
   |--webui
   |--amf.zip
   |--hss.zip
   |--mongodb.zip
   |--pcrf.zip
   |--smf.zip
   |--upf.zip
   |--webui.zip

Create and upload VNF packages

Under kube5gnfvo/example/free5gcv1
  • Create VNF packages We send request to API server in "POST" method.
  1. You can modify following URL, replace as your IP address, and fill it in the request URL column.
http://<your server IP>:30888/vnfpkgm/v1/vnf_packages/

Send request to server, you will receive server response "201 Created".

  • Upload VNF packages On the previous step, server will respone an id number(see the figure shown below, the red box)
  1. Create a new request in "PUT" method. Modify following URL, replace as your IP address; as we got in last figure.
http://<your server IP>:30888/vnfpkgm/v1/vnf_packages/<id>/package_content/

  1. Modify the headers of the request. Add two key-value, "Accept-application/zip" and "Accept-application/json" as following figure shown.

  2. Modify the body of the request. In "form-data" format, we add one key-value: "file" is our key and seven .zip files we compressed under /vnfpackages directory is our "value". Remenber to select value type as "file" then you can choose .zip file as your value. To select value type you can find the menu as shown in figure(red box). In each request we will upload only "one" file.

  1. Send the request to server, you will get response "202 Accepted".

  2. Repeat step 1 to step 4, each time replace one .zip file under /vnfpackages directory. Not until upload seven .zip file will you go to next step (step 7).

Create ns descriptor

  1. Create a new request in "POST" method. Modify following URL, replace as your IP address.
http://<your server IP>:30888/nsd/v1/ns_descriptors/

Send request to server. You will receive server response "201 Created", and an id number as shown in following figure (red box).

Create and upload ns descriptor

  1. Create a new request in "PUT" method. Modify following URL, replace as your IP address; as we got in last figure.
http://<your server IP>:30888/nsd/v1/ns_descriptors/<id>/nsd_content/

  1. Modify the headers of the request. Add two key-value, "Accept-application/zip" and "Accept-application/json" as following figure shown.

  2. Modify the body of the request. In "form-data" format, we add one key-value: "file" is our key and "value" is our ns.zip file we comressed before. Also remenber to select type of value as "file"(shown in following figure, red box). Send the request to server, you will receive server response "202 Accepted".

Create and instantiate ns instance

  1. Create a new request in "POST" method. Modify following URL, replace as your IP address.
http://<your server IP>:30888/nslcm/v1/ns_instances/
  1. Modify the body of the request. In "raw, JSON" format, type in following JSON file.
{
    "nsdId": "2116fd24-83f2-416b-bf3c-ca1964793acb",
    "nsName": "String",
    "nsDescription": "String"
}

Send the request to server, you will receive server response "201 Created", and body of response in JSON format. Retain the response, we will use them in next step.

  1. Create a new request in "POST" method. Modify following URL, replace as your IP address; as we got in last response(as following figure shown, in red box).
http://<your server IP>:30888/nslcm/v1/ns_instances/<id in red box>/instantiate/

  1. In request body, we used "raw, JSON" format.You can fill in the body as following.
{
    "vnfInstanceData":[
        {
            "vnfInstanceId": "<id in second layer>",
            "vnfProflieId": "String"
        },
        {
            "vnfInstanceId": "<id in second layer>",
            "vnfProflieId": "String"
        },
        {
            "vnfInstanceId": "<id in second layer>",
            "vnfProflieId": "String"
        },
        {
            "vnfInstanceId": "<id in second layer>",
            "vnfProflieId": "String"
        },
        {
            "vnfInstanceId": "<id in second layer>",
            "vnfProflieId": "String"
        },
        {
            "vnfInstanceId": "<id in second layer>",
            "vnfProflieId": "String"
        },
        {
            "vnfInstanceId": "<id in second layer>",
            "vnfProflieId": "String"
        }
    ]
}

For "vnfInstanceId" fields, you can get them in previous response. The response in JSON format, we used ID in second layer as vnfInstanceId. The following is the JSON structure of response to show how we use in vnfInstanceId. You can also see the blue box in previous figure as example.

{
    "id": "ce76d430-f8d8-4598-9229-cdae44ab9f32",
    "vnffgInfo": [],
    "vnfInstance": [
        {
            "id": "<vnfInstanceId we look for>",
            "vnfInstanceName": "2116fd24-83f2-416b-bf3c-ca1964793smf-iqfuvhakgw",
            ......
        },
        {
            "id": "<vnfInstanceId we look for>",
            "vnfInstanceName": "2116fd24-83f2-416b-bf3c-ca19647webui-augxvnqoyw",
            ......
        },
        ......
}

Send the request to server, you will receive server response "202 Accepted".