Skip to main content

Kubernetes Overview

What is Kubernetes?​

Kubernetes is an open-source container orchestration platform that automates the deployment, scaling, and management of containerized applications. Key points include:

  • Developed originally by Google, now maintained by the Cloud Native Computing Foundation
  • Designed to run distributed systems resiliently
  • Provides features like service discovery, load balancing, storage orchestration, and self-healing
  • Enables declarative configuration and automation

Kubernetes Architecture​

Kubernetes uses a master-worker architecture:

  1. Master Node (Control Plane):

    • API Server: Central management point for the cluster (this is the frontend for the cluster)
    • Scheduler: Assigns work to nodes
    • Controller Manager: Regulates the state of the system
    • Cloud Controller Manager: Regulates the state of the system, in case we are using a cloud provider
    • etcd: Distributed key-value store for cluster data (mini database)

  2. Worker Nodes:

    • Kubelet: Ensures containers are running in a pod
    • Container Runtime: Software for running containers (e.g., Docker)
    • Kube-proxy: Manages network rules on nodes

cluster

Other architectures​

Orchestrated architecture​

orchestrated-architecture

Key Kubernetes Concepts​

  1. Clusters:

    • A set of nodes that run containerized applications
    • Provides high availability and scalability (you can have replicas in different datacenters)

  2. Nodes:

    • Physical or virtual machines in the Kubernetes cluster
    • Can be master nodes (part of the control plane) or worker nodes

  3. Namespaces:

    • Virtual clusters within a physical cluster
    • Used for organizing resources and multi-tenancy

  4. Pods:

    • Smallest deployable units in Kubernetes
    • Can contain one or more containers
    • Share network namespace and storage

  5. Replica Sets:

    • Describe the desired state for a set of pods
    • Provide basic scaling and self-healing mechanisms

  6. Services:

    • An abstract way to expose applications running on pods
    • Provide a stable network endpoint

  7. Persistent Volumes (and Persistent Volume Claims):

    • Abstraction for storage resources in the cluster

  8. ConfigMaps and Secrets:

    • Manage configuration data and sensitive information

  9. Ingress:

    • Manage external access to services in a cluster