Local installation
Prerequisites
Make sure the following tools are installed in the environment before installing KubeAdmiral:
Installation steps
If you want to understand how KubeAdmiral works, you can easily start a cluster with KubeAdmiral control plane on your local computer.
1.Clone the KubeAdmiral repository to your local environment：
$ git clone https://github.com/kubewharf/kubeadmiral
2.Switch to the KubeAdmiral directory:
$ cd kubeadmiral
3.Install and start KubeAdmiral:
$ make local-up
The command performs the following tasks mainly:
- Use the Kind tool to start a meta-cluster;
- Install the KubeAdmiral control plane components on the meta-cluster;
- Use the Kind tool to start 3 member clusters;
- Bootstrap the joining of the 3 member clusters to the federated control plane;
- Export the cluster kubeconfigs to the $HOME/.kube/kubeadmiral directory.
If all the previous steps went successfully, we would see the following message:
Your local KubeAdmiral has been deployed successfully!
To start using your KubeAdmiral, run:
export KUBECONFIG=$HOME/.kube/kubeadmiral/kubeadmiral.config
To observe the status of KubeAdmiral control-plane components, run:
export KUBECONFIG=$HOME/.kube/kubeadmiral/meta.config
To inspect your member clusters, run one of the following:
export KUBECONFIG=$HOME/.kube/kubeadmiral/member-1.config
export KUBECONFIG=$HOME/.kube/kubeadmiral/member-2.config
export KUBECONFIG=$HOME/.kube/kubeadmiral/member-3.config
4.Wait for all member clusters to join KubeAdmiral
After the member clusters have successfully joined KubeAdmiral, we can observe the detailed status of the member clusters through the control plane: three member clusters have joined KubeAdmiral, and their statuses are all READY and JOINED.
$ kubectl get fcluster
NAME READY JOINED AGE
kubeadmiral-member-1 True True 1m
kubeadmiral-member-2 True True 1m
kubeadmiral-member-3 True True 1m
Installation KubeAdmiral by Helm Chart
Prerequisites
Make sure the following tools are installed in the environment before installing KubeAdmiral:
Installation steps
If you already have a Kubernetes cluster, you can install the KubeAdmiral control plane on your cluster using the helm chart. To install KubeAdmiral, follow these steps:
1.Get the Chart package for KubeAdmiral and install it:
Get the Chart package locally and install it.
$ git clone https://github.com/kubewharf/kubeadmiral
$ cd kubeadmiral
$ helm install kubeadmiral -n kubeadmiral-system --create-namespace --dependency-update ./charts/kubeadmiral
2.Wait and check if the package has been installed successfully
Use your Kubernetes cluster kubeconfig to see if the following components of KubeAdmiral have been successfully running:
$ kubectl get pods -n kubeadmiral-system
NAME READY STATUS RESTARTS AGE
etcd-0 1/1 Running 0 13h
kubeadmiral-apiserver-5767cd4f56-gvnqq 1/1 Running 0 13h
kubeadmiral-controller-manager-5f598574c9-zjmf9 1/1 Running 0 13h
kubeadmiral-hpa-aggregator-59ccd7b484-phbr6 2/2 Running 0 13h
kubeadmiral-kube-controller-manager-6bd7dcf67-2zpqw 1/1 Running 2 (13h ago) 13h
3.Export the kubeconfig of KubeAdmiral
After executing the following command, the kubeconfig for connecting to KubeAdmiral will be exported to the kubeadmiral-kubeconfig file.
Note that the address in the kubeconfig is set to the internal service address of KubeAdmiral-apiserver:
$ kubectl get secret -n kubeadmiral-system kubeadmiral-kubeconfig-secret -o jsonpath={.data.kubeconfig} | base64 -d > kubeadmiral-kubeconfig
If you specified an external address when installing KubeAdmiral, we will automatically generate a kubeconfig using the external address. You can export it to the external-kubeadmiral-kubeconfig file by running the following command:
$ kubectl get secret -n kubeadmiral-system kubeadmiral-kubeconfig-secret -o jsonpath={.data.external-kubeconfig} | base64 -d > external-kubeadmiral-kubeconfig
Uninstallation steps
Uninstall the KubeAdmiral Helm chart in the kubeadmiral-system namespace:
$ helm uninstall -n kubeadmiral-system kubeadmiral
This command will delete all Kubernetes resources associated with the Chart:
Note: The following permissions and namespace resources are relied on when installing and uninstalling helmchart, so they cannot be deleted automatically and require you to clean them up manually.
$ kubectl delete clusterrole kubeadmiral-pre-install-job
$ kubectl delete clusterrolebinding kubeadmiral-pre-install-job
$ kubectl delete ns kubeadmiral-system
Configuration parameters
| Name | Description | Default Value |
|---|---|---|
| clusterDomain | Default cluster domain of Kubernetes cluster | “cluster.local” |
| etcd.image.name | Image name used by KubeAdmiral etcd | “registry.k8s.io/etcd:3.4.13-0” |
| etcd.image.pullPolicy | Pull mode of etcd image | “IfNotPresent” |
| etcd.certHosts | Hosts accessible with etcd certificate | [“kubernetes.default.svc”, “.etcd.{{ .Release.Namespace }}.svc.{{ .Values.clusterDomain }}”, “.{{ .Release.Namespace }}.svc.{{ .Values.clusterDomain }}”, “.{{ .Release.Namespace }}.svc”, “localhost”, “127.0.0.1”] |
| apiServer.image.name | Image name of kubeadmiral-apiserver | “registry.k8s.io/kube-apiserver:v1.20.15” |
| apiServer.image.pullPolicy | Pull mode of kubeadmiral-apiserver image | “IfNotPresent” |
| apiServer.certHosts | Hosts supported by kubeadmiral-apiserver certificate | [“kubernetes.default.svc”, “.etcd.{{ .Release.Namespace }}.svc.{{ .Values.clusterDomain }}”, “.{{ .Release.Namespace }}.svc.{{ .Values.clusterDomain }}”, “.{{ .Release.Namespace }}.svc”, “localhost”, “127.0.0.1”] |
| apiServer.hostNetwork | Deploy kubeadmiral-apiserver with hostNetwork. If there are multiple kubeadmirals in one cluster, you’d better set it to “false” | “false” |
| apiServer.serviceType | Service type of kubeadmiral-apiserver | “ClusterIP” |
| apiServer.externalIP | Exposed IP of kubeadmiral-apiserver. If you want to expose the apiserver to the outside, you can set this field, which will write the external IP into the certificate and generate a kubeconfig with the external IP. | "" |
| apiServer.nodePort | Node port used for the ‘apiserver’. This will take effect when ‘apiServer.serviceType’ is set to ‘NodePort’. If no port is specified, a node port will be automatically assigned. | 0 |
| apiServer.certHosts | Hosts supported by the kubeadmiral-apiserver certificate | [“kubernetes.default.svc”, “.etcd.{{ .Release.Namespace }}.svc.{{ .Values.clusterDomain }}”, “.{{ .Release.Namespace }}.svc.{{ .Values.clusterDomain }}”, “.{{ .Release.Namespace }}.svc”, “localhost”, “127.0.0.1”, “{{ .Values.apiServer.externalIP }}”] |
| kubeControllerManager.image.name | Image name of kube-controller-manager | “registry.k8s.io/kube-controller-manager:v1.20.15” |
| kubeControllerManager.image.pullPolicy | Pull mode of kube-controller-manager image | “IfNotPresent” |
| kubeControllerManager.controllers | Controllers that kube-controller-manager component needs to start | “namespace,garbagecollector” |
| kubeadmiralControllerManager.image.name | Image name of kubeadmiral-controller-manager | “docker.io/kubewharf/kubeadmiral-controller-manager:v1.0.0” |
| kubeadmiralControllerManager.image.pullPolicy | Pull mode of kubeadmiral-controller-manager image | “IfNotPresent” |
| kubeadmiralControllerManager.extraCommandArgs | Additional startup parameters of kubeadmiral-controller-manager | {} |
| kubeadmiralHpaAggregator.image.name | Image name of kubeadmiral-hpa-aggregator | “docker.io/kubewharf/kubeadmiral-hpa-aggregator:v1.0.0” |
| kubeadmiralHpaAggregator.image.pullPolicy | Pull mode of kubeadmiral-hpa-aggregator image | “IfNotPresent” |
| kubeadmiralHpaAggregator.extraCommandArgs | Additional startup parameters of kubeadmiral-hpa-aggregator | {} |
| installTools.cfssl.image.name | cfssl image name for KubeAdmiral installer | “docker.io/cfssl/cfssl:latest” |
| installTools.cfssl.image.pullPolicy | cfssl image pull policy | “IfNotPresent” |
| installTools.kubectl.image.name | kubectl image name for KubeAdmiral installer | “docker.io/bitnami/kubectl:1.22.10” |
| installTools.kubectl.image.pullPolicy | kubectl image pull policy | “IfNotPresent” |