Kubernetes

Updated on 2025-06-26 GMT+08:00

View PDF

What Is Kubernetes?

Kubernetes is a containerized application software system that can be easily deployed and managed. It facilitates container scheduling and orchestration.

For application developers, Kubernetes can be regarded as a cluster operating system. Kubernetes provides functions such as service discovery, scaling, load balancing, self-healing, and even leader election, freeing developers from infrastructure-related configurations.

When you use Kubernetes, it is like running a giant number of servers all on the same physical machine, and then running your applications on that single massive platform. Kubernetes enables you to deploy applications always using the same method, regardless of how many servers there are in a cluster.

Figure 1 Running applications in a Kubernetes cluster

Kubernetes Cluster Architecture

A Kubernetes cluster consists of master nodes and worker nodes. Applications are deployed on worker nodes, and you can specify the nodes for deployment.

NOTE:

For CCE clusters, master nodes are managed by CCE. You only need to create worker nodes.

The following figure shows the architecture of a Kubernetes cluster.

Figure 2 Kubernetes cluster architecture

Master node

A master node is the machine where the control plane components run, including the API server, scheduler, controller manager, and etcd.

API server: a transit station for components to communicate with each other. It receives external requests and writes data to etcd.
Controller manager: carries out cluster-level functions, such as component replication, worker node tracing, and node fault troubleshooting.
Scheduler: schedules pods to nodes based on various conditions (such as available resources and node affinity).
etcd: provides distributed data storage for cluster configurations.

In a production environment, multiple master nodes are deployed to ensure high cluster availability. For example, you can deploy three master nodes in a CCE cluster.

Worker node

A worker node is a compute node for containerized applications in a cluster. A worker node consists of the following components:

kubelet: communicates with the container runtime, interacts with the API server, and manages pods on the node.
kube-proxy: serves as an access proxy between application components.
Container runtime: functions as an engine such as Docker for downloading images and running containers.

Kubernetes Scalability

Kubernetes makes the Container Runtime Interface (CRI), Container Network Interface (CNI), and Container Storage Interface (CSI) open-sourced. These interfaces maximize Kubernetes scalability and allow Kubernetes to focus on container scheduling.

CRI: provides computing resources for a container runtime. It makes differences between container engines irrelevant and interacts with each container engine through a unified interface.
CNI: enables Kubernetes to support different networking implementations. For example, the custom CNI add-on of CCE allows your Kubernetes clusters to run in VPCs.
CSI: enables Kubernetes to support various classes of storage. For example, CCE can be interconnected with block storage (EVS), file storage (SFS), and object storage (OBS) services.

Basic Objects in Kubernetes

The following figure shows the basic objects in Kubernetes and the relationships between them.

Figure 3 Basic Kubernetes objects

Pod
Pods are the smallest deployable units of compute that you can create and manage in Kubernetes. A pod is a group of one or more containers. Pods have shared storage, unique IP addresses, and specifications for how to run containers.
Deployment
A Deployment is a service-oriented encapsulation of pods. It can contain one or more pods. These pods have the same role, and the system automatically distributes requests to the pods of a Deployment.
StatefulSet
A StatefulSet is used to manage stateful applications. Like a Deployment, a StatefulSet manages a group of pods that are based on an identical container specification. Unlike a Deployment, a StatefulSet maintains a fixed ID for each of their pods. These pods are created based on the same declaration, but they are not interchangeable. Each pod has a permanent ID regardless of how it was scheduled.
Job
A job is used to control batch tasks. Jobs are different from long-term servo tasks (such as Deployments). Jobs are started and terminated at specific times, but long-term servo tasks continue to run until they are specifically terminated. The pods managed by a job will be automatically removed after successfully completing tasks based on user configurations.
CronJob
A CronJob is a time-based job. Similar to the crontab of Linux, it runs a specified job in a specified time range.
DaemonSet
A DaemonSet runs only one pod on each node in a cluster. This works well for certain system-level applications such as log collection and resource monitoring since they must run on each node and need only a few pods. A good example is kube-proxy.
Service
A Service is used for pod access. With a fixed IP address, a Service forwards access traffic to pods and balances load for these pods.
Ingress
Services forward requests at Layer 4 using TCP and UDP. Ingresses can forward requests at Layer 7 using HTTP and HTTPS and make forwarding more targeted by domain names and paths.
ConfigMap
A ConfigMap stores configurations in key-value pairs required by applications. ConfigMaps allow you to decouple environment-specific configurations from container images, so that different environments can have their own unique configurations.
Secret
A secret lets you store and manage sensitive information, such as authentication information, certificates, and private keys. Storing confidential information in a secret is safer and more flexible than putting it verbatim in a pod definition or in a container image.
PersistentVolume (PV)
A PV describes a persistent data storage volume. It defines a directory for persistent storage on a host machine, for example, a mount directory of a network file system (NFS).
PersistentVolumeClaim (PVC)
A PVC in Kubernetes is a request for persistent storage. PVCs free you from creating and releasing underlying storage resources. With PVCs, you only need to specify the storage type and capacity.

Setting Up a Kubernetes Cluster

Kubernetes introduces multiple methods for setting up a Kubernetes cluster, such as minikube and kubeadm.

If you do not want to set up a Kubernetes cluster by yourself, you can purchase one on the CCE console. The following uses clusters purchased on the CCE console as examples.

Kubernetes Objects

Resources in Kubernetes can be described in YAML or JSON format. An object consists of the following parts:

typeMeta: metadata of the object type. It specifies the API version and type of the object.
objectMeta: metadata of the object, such as the object name and labels.
spec: desired status of the object, for example, which image the object uses and how many replicas the object has.
status: actual status of the object. It can only be viewed after the object is created. You do not need to specify the status when creating an object.

Figure 4 YAML file

Running Applications on Kubernetes

Delete status from the content in Figure 4 and save it as the nginx-deployment.yaml file, as shown below:

apiVersion: apps/v1
kind: Deployment
metadata:
  name:  nginx
  labels:
    app:  nginx
spec:
  selector:
    matchLabels:
      app: nginx
  replicas: 3
  template:
    metadata:
      labels:
    app:  nginx
    spec:
      containers:
      - name:  nginx
        image:  nginx:alpine
        resources:
          requests:
            cpu: 100m
            memory: 200Mi
          limits:
            cpu: 100m
            memory: 200Mi
      imagePullSecrets:
      - name: default-secret

Use kubectl to access the cluster and run the following command:

# kubectl create -f nginx-deployment.yaml 
deployment.apps/nginx created

After the command is executed, three pods are created in the Kubernetes cluster. You can run the following command to obtain the Deployment and pods:

# kubectl get deploy
NAME    READY   UP-TO-DATE   AVAILABLE   AGE
nginx   3/3     3            3           9s

# kubectl get pods
NAME                     READY   STATUS    RESTARTS   AGE
nginx-685898579b-qrt4d   1/1     Running   0          15s
nginx-685898579b-t9zd2   1/1     Running   0          15s
nginx-685898579b-w59jn   1/1     Running   0          15s

So far, we have walked you through containers and Docker, Kubernetes cluster setup, and basic Kubernetes concepts, and provided you an example of how to use kubectl. The following sections will go deeper into Kubernetes objects and how they are used and related.

Parent topic: Containers and Kubernetes

Previous topic: Containers

Next topic: Using kubectl to Operate a Cluster