Using Kubernetes

4 Using Kubernetes

Important:

The software described in this documentation is either in Extended Support or Sustaining Support. See Oracle Open Source Support Policies for more information.

We recommend that you upgrade the software described by this documentation as soon as possible.

This chapter describes how to get started using Kubernetes to deploy, maintain and scale your containerized applications. In this chapter, we describe basic usage of the kubectl command to get you started creating and managing containers and services within your environment.

The kubectl utility is fully documented in the upstream documentation at:

https://kubernetes.io/docs/reference/kubectl/

About Runtime Engines

runc is the default runtime engine when you create containers. You can also use the kata-runtime runtime engine to create Kata containers. For information on Kata containers and how to create them, see Container Runtimes.

Getting Information about Nodes

To get a listing of all of the nodes in a cluster and the status of each node, use the kubectl get command. This command can be used to obtain listings of any kind of resource that Kubernetes supports. In this case, the nodes resource:

kubectl get nodes
NAME                   STATUS    ROLES   AGE      VERSION
control.example.com    Ready     master  1h       v1.21.x+x.x.x.el8
worker1.example.com    Ready     <none>  1h       v1.21.x+x.x.x.el8
worker2.example.com    Ready     <none>  1h       v1.21.x+x.x.x.el8

You can get more detailed information about any resource using the kubectl describe command. If you specify the name of the resource, the output is limited to information about that resource alone; otherwise, full details of all resources are also printed to screen. For example:

kubectl describe nodes worker1.example.com
Name:               worker1.example.com
Roles:              <none>
Labels:             beta.kubernetes.io/arch=amd64
                    beta.kubernetes.io/os=linux
                    kubernetes.io/arch=amd64
                    kubernetes.io/hostname=worker1.example.com
                    kubernetes.io/os=linux
Annotations:        flannel.alpha.coreos.com/backend-data: {"VtepMAC":"fe:78:5f:ea:7c:c0"}
                    flannel.alpha.coreos.com/backend-type: vxlan
                    flannel.alpha.coreos.com/kube-subnet-manager: true
                    flannel.alpha.coreos.com/public-ip: 192.0.2.11
                    kubeadm.alpha.kubernetes.io/cri-socket: /var/run/crio/crio.sock
                    node.alpha.kubernetes.io/ttl: 0
                    volumes.kubernetes.io/controller-managed-attach-detach: true
...

Running an Application in a Pod

To create a pod with a single running container, you can use the kubectl create command. For example:

kubectl create deployment --image nginx hello-world
deployment.apps/hello-world created

Substitute nginx with a container image. Substitute hello-world with a name for your deployment. Your pods are named by using the deployment name as a prefix.

Tip:

Deployment, pod and service names conform to a requirement to match a DNS-1123 label. These must consist of lower case alphanumeric characters or -, and must start and end with an alphanumeric character. The regular expression that is used to validate names is '[a-z0-9]([-a-z0-9]*[a-z0-9])?'. If you use a name for your deployment that does not validate, an error is returned.

There are many additional optional parameters that can be used when you run a new application within Kubernetes. For instance, at run time, you can specify how many replica pods should be started, or you might apply a label to the deployment to make it easier to identify pod components. To see a full list of options available to you, run kubectl run --help.

To check that your new application deployment has created one or more pods, use the kubectl get pods command:

kubectl get pods
NAME                           READY     STATUS    RESTARTS   AGE
hello-world-5f55779987-wd857   1/1       Running   0          1m

Use kubectl describe to show a more detailed view of your pods, including which containers are running and what image they are based on, as well as which node is currently hosting the pod:

kubectl describe pods
Name:               hello-world-5f55779987-wd857
Namespace:          default
Priority:           0
PriorityClassName:  <none>
Node:               worker1.example.com/192.0.2.11
Start Time:         Fri, 16 Aug 2019 08:48:33 +0100
Labels:             app=hello-world
                    pod-template-hash=5f55779987
Annotations:        <none>
Status:             Running
IP:                 10.244.1.3
Controlled By:      ReplicaSet/hello-world-5f55779987
Containers:
  nginx:
    Container ID:   cri-o://417b4b59f7005eb4b1754a1627e01f957e931c0cf24f1780cd94fa9949be1d31
    Image:          nginx
    Image ID:       docker-pullable://nginx@sha256:5d32f60db294b5deb55d078cd4feb410ad88e6fe7...
    Port:           <none>
    Host Port:      <none>
    State:          Running
      Started:      Mon, 10 Dec 2018 08:25:25 -0800
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-s8wj4 (ro)
Conditions:
  Type              Status   
  Initialized       True
  Ready             True
  ContainersReady   True
  PodScheduled      True
Volumes:
  default-token-s8wj4:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-s8wj4
    Optional:    false
QoS Class:       BestEffort  
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
....

Scaling a Pod Deployment

To change the number of instances of the same pod that you are running, you can use the kubectl scale deployment command. For example:

kubectl scale deployment --replicas=3 hello-world
deployment.apps/hello-world scaled

You can check that the number of pod instances has been scaled appropriately:

kubectl get pods
NAME                           READY     STATUS    RESTARTS   AGE
hello-world-5f55779987-tswmg   1/1       Running   0          18s
hello-world-5f55779987-v8w5h   1/1       Running   0          26m
hello-world-5f55779987-wd857   1/1       Running   0          18s

Exposing a Service Object for an Application

Typically, while many applications only need to communicate internally within a pod, or even across pods, you might need to expose your application externally so that clients outside of the Kubernetes cluster can interface with the application. You can do this by creating a service definition for the deployment.

Note:

Prerequisite for LoadBalancer service:

The Oracle Cloud Infrastructure Cloud Controller Manager module (oci-ccm) is used to create and manage Oracle Cloud Infrastructure load balancers for Kubernetes applications. Hence, the following example assumes you have completed installation of the oci-ccm module as described in Application Load Balancers.

To expose a deployment using a service object, you must define the service type that should be used. The following example shows how you might use the kubectl expose deployment command to expose the application via a LoadBalancer service:

kubectl expose deployment hello-world --port 80 --type=LoadBalancer
service/hello-world exposed

Use kubectl get services to list the different services that the cluster is running as shown in the following example. Note that the EXTERNAL-IP field of the LoadBalancer service initially shows as <pending> whilst the setup of the service is still in progress:

kubectl get services
NAME          TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
hello-world   LoadBalancer   10.102.42.160   <pending>     80:31847/TCP   3s
kubernetes    ClusterIP      10.96.0.1       <none>        443/TCP        5h13m

You can see the load balancer in the Oracle Cloud Infrastructure console. Initially, its state in the console is shown as Creating.

Wait a few minutes for the setup of the service to complete. Run the kubectl get services command again, and note that the EXTERNAL-IP field is now populated with the IP address assigned to the LoadBalancer service:

kubectl get services
NAME          TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
hello-world   LoadBalancer   10.102.42.160   192.0.2.250   80:31847/TCP   85s
kubernetes    ClusterIP      10.96.0.1       <none>        443/TCP        5h15m

In the preceding sample output the PORT(s) field contains the following ports:

Port 80: The port at which the LoadBalancer service can be accessed. In this example, the service would be accessed at the following URL:
```
http://192.0.2.250
```
Port 31847: The port assigned to the NodePort service. The NodePort service enables the application to be accessed using URL format worker_node:NodePort, for example:
```
http://worker1.example.com:31847/
```
Note:
Kubernetes creates the NodePort service as part of its LoadBalancer setup.

You can verify the services have been set up successfully by running curl commands as shown in the following examples:

For the LoadBalancer service:

curl http://192.0.2.250

    <html>
        <head>
            <title>Welcome to this servicer</title>
        </head>
        <body>
                <h1>Welcome to this service</h1>
                ...
        </body>
    </html>

For each worker node, verify the NodePort service by running a curl command as illustrated below:

curl http://worker1.example.com:31847/

     <html>
        <head>
            <title>Welcome to this servicer</title>
            ...
        </head>
        <body>
                <h1>Welcome to this service</h1>
                ...
        </body>
    </html>

Deleting a Service or Deployment

Objects can be deleted easily within Kubernetes so that your environment can be cleaned. Use the kubectl delete command to remove an object.

To delete a service, specify the services object and the name of the service that you want to remove. For example:

kubectl delete services hello-world
service "hello-world" deleted

To delete an entire deployment, and all of the pod replicas running for that deployment, specify the deployment object and the name that you used to create the deployment:

kubectl delete deployment hello-world
deployment.extensions "hello-world" deleted

Working With Namespaces

Namespaces can be used to further separate resource usage and to provide limited environments for particular use cases. By default, Kubernetes configures a namespace for Kubernetes system components and a standard namespace to be used for all other deployments for which no namespace is defined.

To view existing namespaces, use the kubectl get namespaces and kubectl describe namespaces commands.

The kubectl command only displays resources in the default namespace, unless you set the namespace specifically for a request. Therefore, if you need to view the pods specific to the Kubernetes system, you would use the --namespace option to set the namespace to kube-system for the request. For example, in a cluster with a single control plane node:

kubectl get pods --namespace=kube-system
NAME                                    READY   STATUS    RESTARTS   AGE
coredns-5bc65d7f4b-qzfcc                1/1     Running   0          23h
coredns-5bc65d7f4b-z64f2                1/1     Running   0          23h
etcd-control1.example.com               1/1     Running   0          23h
kube-apiserver-control1.example.com     1/1     Running   0          23h
kube-controller-control1.example.com    1/1     Running   0          23h
kube-flannel-ds-2sjbx                   1/1     Running   0          23h
kube-flannel-ds-njg9r                   1/1     Running   0          23h
kube-proxy-m2rt2                        1/1     Running   0          23h
kube-proxy-tbkxd                        1/1     Running   0          23h
kube-scheduler-control1.example.com     1/1     Running   0          23h
kubernetes-dashboard-7646bf6898-d6x2m   1/1     Running   0          23h

Using Deployment Files

To simplify the creation of pods and their related requirements, you can create a deployment file that define all of the elements that comprise the deployment. This deployment defines which images should be used to generate the containers within the pod, along with any runtime requirements, as well as Kubernetes networking and storage requirements in the form of services that should be configured and volumes that may need to be mounted.

Deployments are described in detail at:

https://kubernetes.io/docs/concepts/workloads/controllers/deployment/