Kubernetes Architecture

This section explains the architecture and key components of Kubernetes, focusing on how the control plane and worker nodes operate together.

  • Control Plane Components:

    • kube-apiserver
    • etcd
    • kube-scheduler
    • kube-controller-manager
    • cloud-controller-manager

  • Worker Node Components:

    • kubelet
    • kube-proxy
    • Container Runtime

Note: The diagram below represents the architecture of Kubernetes clusters and components.

alt text

Control Plane

  • kube-apiserver:

    • The API server is the entry point for all administrative tasks in a Kubernetes cluster. It handles RESTful API requests from users and cluster components.
    • It performs authentication, authorization, and resource management by interfacing with etcd.

  • etcd:

    • A highly available and distributed key-value store that stores all cluster data, including the configuration, state, and metadata of Kubernetes objects like pods, services, etc.
    • It ensures that any update made to the cluster is stored and accessible across all control plane components.

  • kube-scheduler:

    • The scheduler is responsible for assigning new pods to available nodes. It considers various constraints, like resource requirements and policies, to ensure pods are efficiently placed.

  • kube-controller-manager:

    • This component runs the core control loops that watch for changes in cluster state. If the desired state does not match the actual state, it takes corrective action.
    • It manages built-in controllers like Deployment, ReplicaSet, and Job.

  • cloud-controller-manager:

    • This controller manages integration with cloud provider APIs (e.g., AWS, GCP). It ensures that resources like load balancers and storage are provisioned based on cloud-specific requirements.

Worker Nodes

  • kubelet:

    • The kubelet is the agent that runs on each worker node. It ensures containers are running in pods and reports the node and pod status to the control plane.
    • It interacts with the container runtime to manage containers.

  • kube-proxy:

    • Kube-proxy runs on every worker node to manage network rules and ensure proper routing of traffic to pods.
    • It supports communication between different services within the cluster and external traffic.

  • Container Runtime:

    • The container runtime, such as containerd or Docker, is responsible for pulling container images and managing the container lifecycle within pods.

Architecture Flow

  • User Request: When a user interacts with Kubernetes (e.g., deploying an application), they send a request to the kube-apiserver using kubectl.
  • API Server Interaction: The API server processes the request and records changes in etcd. If a new pod needs to be created, the API server passes this information to the scheduler.
  • Scheduler Action: The scheduler selects a suitable worker node and assigns the pod to it.
  • Node Operations: The kubelet on the selected worker node pulls the container image using the container runtime, starts the pod, and continuously monitors its health.
  • Networking: Kube-proxy manages the network rules to ensure the pod is accessible based on the assigned service.

References