This page is showing how to setup Free5GC on a KinD cluster hosted in a Linux virtual machine. KinD has some specific configuration to be considered when deploying Free5GC.
As documented in the Free5GC documentation:
- Virtual Machine shall be installed (at least 6GB RAM, 40 GB DISK), e.g. lubuntu 22.04 LTS.
- Git clone the towards5gs-helm project
- Kind, kubectl, helm commands to be installed
- See related web sites. I copy the binaries in /usr/local/bin/ and need to use sudo to execute the commands.
- GTP5G kernel module to be installed on the host
git clone https://github.com/free5gc/gtp5g.git && cd gtp5g make clean & make sudo make install - Creation of a KinD cluster: I choose to use one node for control plane and one worker node.
- I edit the following configuration file:
kind: Cluster apiVersion: kind.x-k8s.io/v1alpha4 nodes: - role: control-plane - role: worker - I create the KinD cluster with the following command
cub@labFormation:~$ sudo kind create cluster --config mycluster.yaml Creating cluster "kind" ... ✓ Ensuring node image (kindest/node:v1.27.3) 🖼 ✓ Preparing nodes 📦 📦 ✓ Writing configuration 📜 ✓ Starting control-plane 🕹 ✓ Installing CNI 🔌 ✓ Installing StorageClass 💾 ✓ Joining worker nodes 🚜 - One issue with KinD is that CNI plugins are not installed in the cluster nodes. We need to add them manually.
- Get the latest binaries package, e.g. https://github.com/containernetworking/plugins/releases/tag/v1.3.0
- Copy the binaries from the archive in the /opt/cni/bin folder of each docker containers executing the KIND cluster. Each node of KinD cluster corresponds to a docker container on the host machine. First we identify them and then we copy the plugin binaries.
cub@labFormation:~/CNI$ sudo docker ps CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES 3013a5ca7d6f kindest/node:v1.27.3 "/usr/local/bin/entr…" 11 minutes ago Up 11 minutes 127.0.0.1:39573->6443/tcp kind-control-plane e3cb8a67b490 kindest/node:v1.27.3 "/usr/local/bin/entr…" 11 minutes ago Up 11 minutes kind-worker cub@labFormation:~/CNI$ sudo docker cp . 3013a5ca7d6f:/opt/cni/bin/ Successfully copied 78.4MB to 3013a5ca7d6f:/opt/cni/bin/ cub@labFormation:~/CNI$ sudo docker cp . e3cb8a67b490:/opt/cni/bin/ Successfully copied 78.4MB to e3cb8a67b490:/opt/cni/bin/
- Install Multus CNI on the KinD cluster. Following the instructions from the multus-cni github repo.:
git clone https://github.com/k8snetworkplumbingwg/multus-cni cub@labFormation:~/multus-cni$ cat ./deployments/multus-daemonset-thick.yml | sudo kubectl apply -f - customresourcedefinition.apiextensions.k8s.io/network-attachment-definitions.k8s.cni.cncf.io created clusterrole.rbac.authorization.k8s.io/multus created clusterrolebinding.rbac.authorization.k8s.io/multus created serviceaccount/multus created configmap/multus-daemon-config created daemonset.apps/kube-multus-ds created - We need to adjust the network configuration of Free5GC related to the KinD cluster. The worker node in the KinD cluster has only one eth0, and we need its address class and GW to configure.
- Running command inside the container executing the worker node:
We got 172.18.0.0/16, assuming the GW is 172.18.0.0root@kind-worker:/# ip a 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 10: eth0@if11: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default link/ether 02:42:ac:12:00:03 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 172.18.0.3/16 brd 172.18.255.255 scope global eth0 valid_lft forever preferred_lft forever inet6 fc00:f853:ccd:e793::3/64 scope global nodad valid_lft forever preferred_lft forever inet6 fe80::42:acff:fe12:3/64 scope link valid_lft forever preferred_lft forever root@kind-worker:/# ip r default via 172.18.0.1 dev eth0 10.244.0.0/24 via 172.18.0.2 dev eth0 172.18.0.0/16 dev eth0 proto kernel scope link src 172.18.0.3 - We need to edit the Free 5GC master helm chart values to use only eth0 as network interface and to add the specific @IP details for the N6 network
n6network: enabled: true name: n6network type: ipvlan masterIf: eth0 subnetIP: 172.18.0.0 cidr: 16 gatewayIP: 172.18.0.0 excludeIP: 172.18.0.0 - We also need to set up an IP@ compatible with the class to the UPF N6 interface. We need to edit the free5gc-upc helm chart value for the corresponding value:
n6if: # DN ipAddress: 172.18.0.22
- I edit the following configuration file:
- Creation of a persistent volume
- From the Free5GC volume example, we need to edit the targeted worker node (here "kind-worker") and path for storage that must be created in the docker container executing the worker node.
- Edition of the template
apiVersion: v1 kind: PersistentVolume metadata: name: example-local-pv9 labels: project: free5gc spec: capacity: storage: 8Gi accessModes: - ReadWriteOnce persistentVolumeReclaimPolicy: Retain local: path: /home/kubedata nodeAffinity: required: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - kind-worker - Creation of folder in the worker node
cub@labFormation:~$ sudo docker exec -it e3cb8a67b490 /bin/bash root@kind-worker:/# mkdir /home/kubedata - Applying the persistent volume creation
cub@labFormation:~$ sudo kubectl apply -f persistent.yaml persistentvolume/example-local-pv9 created
Following Free5GC documentation:
- Creation of namespace (here free5gc)
Creation of the namespace related to free5gc cub@labFormation:~$ sudo kubectl create ns free5gc - Helm install command
cub@labFormation:~/towards5gs-helm/charts$ sudo helm -n free5gc install free5gc-premier ./free5gc/ NAME: free5gc-premier LAST DEPLOYED: Thu Sep 28 17:41:39 2023 NAMESPACE: free5gc STATUS: deployed REVISION: 1 [...] - Verification
cub@labFormation:~/towards5gs-helm/charts$ sudo kubectl get pods -n free5gc NAME READY STATUS RESTARTS AGE free5gc-premier-free5gc-amf-amf-75b48f959d-hd5wd 1/1 Running 0 95s free5gc-premier-free5gc-ausf-ausf-8694b78fff-4zbcw 1/1 Running 0 95s free5gc-premier-free5gc-dbpython-dbpython-7cd595f6b9-xl6w6 1/1 Running 0 95s free5gc-premier-free5gc-nrf-nrf-74579c5897-6wqdq 1/1 Running 0 95s free5gc-premier-free5gc-nssf-nssf-9987c5f49-xc7z5 1/1 Running 0 95s free5gc-premier-free5gc-pcf-pcf-5778596ccf-6ggvb 1/1 Running 0 95s free5gc-premier-free5gc-smf-smf-64d76d858f-8srs7 1/1 Running 0 94s free5gc-premier-free5gc-udm-udm-55975f967f-zsp9p 1/1 Running 0 94s free5gc-premier-free5gc-udr-udr-76976f6779-xv7jh 1/1 Running 0 95s free5gc-premier-free5gc-upf-upf-697d4c85b6-mrv82 1/1 Running 0 95s free5gc-premier-free5gc-webui-webui-5d9c5c9fdb-d98n8 1/1 Running 0 95s mongodb-0- Acccess to Free5GC webUI
WebUI is available through a nodeport, accessible from the host using the external @IP address from the KinD cluster.
- Identification of the nodeport with external port (here 30500)
cub@labFormation:~/towards5gs-helm/charts$ sudo kubectl get svc -n free5gc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE free5gc-premier-free5gc-amf-service ClusterIP 10.96.31.39 <none> 80/TCP 2m27s free5gc-premier-free5gc-ausf-service ClusterIP 10.96.254.209 <none> 80/TCP 2m27s free5gc-premier-free5gc-nssf-service ClusterIP 10.96.33.255 <none> 80/TCP 2m27s free5gc-premier-free5gc-pcf-service ClusterIP 10.96.105.168 <none> 80/TCP 2m27s free5gc-premier-free5gc-smf-service ClusterIP 10.96.53.146 <none> 80/TCP 2m27s free5gc-premier-free5gc-udm-service ClusterIP 10.96.189.56 <none> 80/TCP 2m27s free5gc-premier-free5gc-udr-service ClusterIP 10.96.92.142 <none> 80/TCP 2m27s mongodb ClusterIP 10.96.106.65 <none> 27017/TCP 2m27s nrf-nnrf ClusterIP 10.96.250.83 <none> 8000/TCP 2m27s webui-service NodePort 10.96.239.43 <none> 5000:30500/TCP 2m27s - Identification of the cluster external IP (worker node)
cub@labFormation:~/towards5gs-helm/charts$ sudo kubectl get nodes -o wide NAME STATUS ROLES AGE VERSION INTERNAL-IP EXTERNAL-IP OS-IMAGE KERNEL-VERSION CONTAINER-RUNTIME kind-control-plane Ready control-plane 34m v1.27.3 172.18.0.2 <none> Debian GNU/Linux 11 (bullseye) 6.2.0-33-generic containerd://1.7.1 kind-worker Ready <none> 33m v1.27.3 172.18.0.3 <none> Debian GNU/Linux 11 (bullseye) 6.2.0-33-generic containerd://1.7.1 - Connect to the webUI from the host without proxy: http://172.18.0.3:30500 § Login admin/free5gc § Can add default subscriber via the ui
- Identification of the nodeport with external port (here 30500)
- Acccess to Free5GC webUI
WebUI is available through a nodeport, accessible from the host using the external @IP address from the KinD cluster.
- Use of UERANSIM
cub@labFormation:~/towards5gs-helm/charts$ sudo helm -n free5gc install ueransim-premier ./ueransim/ NAME: ueransim-premier LAST DEPLOYED: Thu Sep 28 17:47:24 2023 NAMESPACE: free5gc STATUS: deployed REVISION: 1 NOTES: # # Software Name : towards5gs-helm # SPDX-FileCopyrightText: Copyright (c) 2021 Orange # SPDX-License-Identifier: Apache-2.0 # # This software is distributed under the Apache License 2.0, # the text of which is available at https://github.com/Orange-OpenSource/towards5gs-helm/blob/main/LICENSE # or see the "LICENSE" file for more details. # # Author: Abderaouf KHICHANE, Ilhem FAJJARI # Software description: An open-source project providing Helm charts to deploy 5G components (Core + RAN) on top of Kubernetes # # # Visit the project at https://github.com/Orange-OpenSource/towards5gs-helm # 1. Run UE connectivity test by running these commands: helm --namespace free5gc test ueransim-premier If you want to run connectivity tests manually, follow: 1. Get the UE Pod name by running: export POD_NAME=$(kubectl get pods --namespace free5gc -l "component=ue" -o jsonpath="{.items[0].metadata.name}") 2. Check that uesimtun0 interface has been created by running these commands: kubectl --namespace free5gc logs $POD_NAME kubectl --namespace free5gc exec -it $POD_NAME -- ip address 3. Try to access internet from the UE by running: kubectl --namespace free5gc exec -it $POD_NAME -- ping -I uesimtun0 www.google.com kubectl --namespace free5gc exec -it $POD_NAME -- curl --interface uesimtun0 www.google.com kubectl --namespace free5gc exec -it $POD_NAME -- traceroute -i uesimtun0 www.google.com Release notes (What's changed in this version): - add the release notes - add an initContainer to wait for the AMF to be ready - enhance the handling of k8s NGAP service and network parameters - Using the command: export POD_NAME=$(sudo kubectl get pods --namespace free5gc -l "component=ue" -o jsonpath="{.items[0].metadata.name}") ○ kubectl --namespace free5gc logs $POD_NAME ○ Showing [2023-09-28 15:47:57.613] [nas] [info] PDU Session establishment is successful PSI[1] [2023-09-28 15:47:57.630] [app] [info] Connection setup for PDU session[1] is successful, TUN interface[uesimtun0, 10.1.0.1] is up.- In the UE:
- In the UE:
As control/worker node are supported by container running in the virtual machine, if local proxy is used, it may not be used from the nodes themselves. Workaround is to use the external @IP of the virtual machine as proxy.
adding configuration of proxies as followed at the end of ~/.bashrc file
export IP_PROXY=$(ifconfig proxy0 | grep "inet " | awk '{print $2}')
export http_proxy=http://${IP_PROXY}:3218/
export https_proxy=http://${IP_PROXY}:3128/
Restart terminal