Kubernetes is now available as a DC/OS package to quickly, and reliably run Kubernetes clusters on Mesosphere DC/OS.
KUBERNETES ON DC/OS IS BETA, DO NOT USE IT FOR PRODUCTION CLUSTERS!
But, please try it out! Give us feedback at: https://github.com/mesosphere/dcos-kubernetes-quickstart/issues
- Google Cloud (GCE) credentials (AWS and Azure are supported as well) with the necessary permissions
- Linux/Mac machine to execute the samples below
- Docker CE 17+
You are now ready to create a DC/OS cluster.
Once the above pre-requisites have been met, clone this repo.
git clone [email protected]:mesosphere/dcos-kubernetes-quickstart.git && cd dcos-kubernetes-quickstart
Set your GCE credentials as environment variables. More information on how to obtain you credentials can be found here
export GOOGLE_APPLICATION_CREDENTIALS=<PATH TO YOUR CREDENTIAL FILE>
The remainder of this quick-start will execute in a Docker container, and create your cluster on GCE, with Kubernetes configured. Simply run
$ make docker
You are now in a container from which you will deploy the cluster and required tools.
$ make deploy
# Installation might take ~ 8minutes.
# Creates a ssh tunnel to a node-agent for APIServer access.
$ make kubectl-tunnel
# Make sure the API Server and Kubelets are up by running:
$ kubectl get nodes
# If you see a result like this, everything is working properly, and you are now running Kubernetes on DC/OS.
NAME STATUS AGE VERSION
kube-node-0-kubelet.kubernetes.mesos Ready 13s v1.7.5
kube-node-1-kubelet.kubernetes.mesos Ready 13s v1.7.5
kube-node-2-kubelet.kubernetes.mesos Ready 13s v1.7.5
make uninstall
# Uninstalls kubernetes.
make destroy-dcos
# Deletes the DC/OS cluster.
The recommended method to install the DC/OS CLI is from the DC/OS web interface. Or, you can manually install the CLI by using the instructions below.
Installing the DC/OS CLI on Linux
Installing the DC/OS CLI on macOS
Use the Kubernetes command-line tool, kubectl, to deploy and manage applications on Kubernetes. Using kubectl, you can inspect cluster resources; create, delete, and update components; and look at your new cluster and bring up example apps.
Follow instructions here to download.
In order to access the Kubernetes API from outside the DC/OS cluster, one needs SSH access to a node-agent. On a terminal window, run:
ssh -N -L 9000:apiserver-insecure.kubernetes.l4lb.thisdcos.directory:9000 <USER>@<HOST>When the Kubernetes API task(s) are healthy, it should be accessible on http://localhost:9000. Reaching this endpoint should show something like this:
$ curl http://localhost:9000
{
"paths": [
"/api",
"/api/v1",
"/apis",
"/apis/apps",
"/apis/apps/v1beta1",
"/apis/authentication.k8s.io",
"/apis/authentication.k8s.io/v1",
"/apis/authentication.k8s.io/v1beta1",
"/apis/authorization.k8s.io",
"/apis/authorization.k8s.io/v1",
"/apis/authorization.k8s.io/v1beta1",
"/apis/autoscaling",
"/apis/autoscaling/v1",
"/apis/autoscaling/v2alpha1",
"/apis/batch",
"/apis/batch/v1",
"/apis/batch/v2alpha1",
"/apis/certificates.k8s.io",
"/apis/certificates.k8s.io/v1beta1",
"/apis/extensions",
"/apis/extensions/v1beta1",
"/apis/policy",
"/apis/policy/v1beta1",
"/apis/rbac.authorization.k8s.io",
"/apis/rbac.authorization.k8s.io/v1alpha1",
"/apis/rbac.authorization.k8s.io/v1beta1",
"/apis/settings.k8s.io",
"/apis/settings.k8s.io/v1alpha1",
"/apis/storage.k8s.io",
"/apis/storage.k8s.io/v1",
"/apis/storage.k8s.io/v1beta1",
"/healthz",
"/healthz/ping",
"/healthz/poststarthook/bootstrap-controller",
"/healthz/poststarthook/ca-registration",
"/healthz/poststarthook/extensions/third-party-resources",
"/logs",
"/metrics",
"/swaggerapi/",
"/ui/",
"/version"
]
}We are now ready to install and configure kubectl, the Kubernetes CLI tool. For the sake of simplicity, we'll be covering the set-up alone:
kubectl config set-cluster dcos-k8s --server=http://localhost:9000
kubectl config set-context dcos-k8s --cluster=dcos-k8s --namespace=default
kubectl config use-context dcos-k8sTest access by retrieving the Kubernetes cluster nodes:
$ kubectl get nodes
NAME STATUS AGE VERSION
kube-node-0-kubelet.kubernetes.mesos Ready 7m v1.7.5
kube-node-1-kubelet.kubernetes.mesos Ready 7m v1.7.5
kube-node-2-kubelet.kubernetes.mesos Ready 7m v1.7.5NOTE: As of this moment, only kube-dns is a mandatory plug-in.
Assuming one has a working Kubernetes cluster with enough available resources, here's how to install the add-on:
kubectl create -f add-ons/dns/kubedns-cm.yaml
kubectl create -f add-ons/dns/kubedns-svc.yaml
kubectl create -f add-ons/dns/kubedns-deployment.yamlNOTE: The Kubernetes namespace where kube-dns will be running is kube-system and not default.
A successful deployment should look like the following:
$ kubectl -n kube-system get deployment,pods
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
deploy/kube-dns 1 1 1 1 1h
NAME READY STATUS RESTARTS AGE
po/kube-dns-1115425399-mwsn8 3/3 Running 0 1hAnd a Kubernetes service should be exposed on 10.100.0.10 with at least one endpoint (pod):
$ kubectl -n kube-system describe svc kube-dns
Name: kube-dns
Namespace: kube-system
Labels: k8s-app=kube-dns
kubernetes.io/cluster-service=true
kubernetes.io/name=KubeDNS
Annotations: <none>
Selector: k8s-app=kube-dns
Type: ClusterIP
IP: 10.100.0.10
Port: dns 53/UDP
Endpoints: 9.0.2.5:53
Port: dns-tcp 53/TCP
Endpoints: 9.0.2.5:53
Session Affinity: None
Events: <none>Here's how to install the Dashboard add-on to the working Kubernetes cluster:
kubectl create -f add-ons/dashboard/kubernetes-dashboard.yamlIf the deployment was successful, point your browser to the url: http://localhost:9000/ui to access the Kubernetes Dashboard.
To deploy your first Kubernetes workloads on DC/OS, please see the examples folder
For more details, please see the docs folder
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent