Kubernetes 101 : Kubectl basic commands

Kubectl command:

Kubectl is the command we use to create, modify delete or check the logs of the kubernetes components (nodes, pods, sevices, replicasets, deployments,...).

Checking the nodes in a cluster:

To check the nodes in a cluster we use the below command:

Deploying pods:

We normally don't deploy pods directly, we go through the deployment "layer":

Example:

We can check the deployment status using:

We get the below result which contains the deployment name and the number of replicas:

To check out the newly deployed pods, we use:

The pod's mame is combination of the deployment_name and a random hash number.

The deployment component takes care of creating the pods.

Replicasets:

Replicasets are used to keep a number of pod replicas to guarantee availability of the applications.

To get the status of replicas, we use:

We can see there are five debian replicas.

Changing the configuration of components:

We get an auto-generated file to which we can make the needed changes.

The old version of the pod terminates and a new one with the new configuration gets deployed.
We can use the "kubectl get pods"to check that.

Checking the logs:

We can check the logs using for debugging purposes:

We can have more information by checking the state of the pod using:

Getting a shell-like access inside the application:

We could use it with nginx or MySQL images for example to get a prompt into the shell of the image.

Deleting a kubernetes component:

To delete a component, for example a pod, we delete the deployment that we used to create it:

We check to see if the pod has been deleted using the "kubectl get pods" command.

Remark:

We create, modify or delete resources at the deployment level, kubernetes takes care of the rest.

Creating components using a Yaml configuration file:

When the amount of options is too overwhelming to fit on the kubectl command line, we can use a Yaml configuration file to create the components. We them run the file using: