Kubernetes is a system that runs and manages containerized applications automatically.

Instead of manually running containers, Kubernetes:

• Starts them
• Stops them
• Restarts them when they fail
• Scales them up or down

• Cluster Fundamentals
  • Cluster
  • Control Plane
  • Node
  • Namespace

• Workloads (Run Applications)
  • Pod
  • ReplicaSet
  • Deployment
  • StatefulSet
  • DaemonSet
  • Job
  • CronJob

• Networking
  • Service
    • ClusterIP
    • NodePort
    • LoadBalancer
    • ExternalName
  • Ingress
  • Ingress Controller
  • NetworkPolicy

• Configuration
  • ConfigMap
  • Secret

• Storage
  • PersistentVolume (PV)
  • PersistentVolumeClaim (PVC)
  • StorageClass
  • Volume

• Security
  • ServiceAccount
  • RBAC
    • Role
    • RoleBinding
    • ClusterRole
    • ClusterRoleBinding
  • SecurityContext
  • Pod Security

• Scheduling
  • NodeSelector
  • Node Affinity
  • Pod Affinity
  • Pod Anti-Affinity
  • Taints
  • Tolerations

• Autoscaling
  • Horizontal Pod Autoscaler (HPA)
  • Vertical Pod Autoscaler (VPA)
  • Cluster Autoscaler

• Observability
  • Logs
  • Metrics
  • Traces
  • Prometheus
  • Grafana
  • Loki
  • Jaeger

• Configuration Management
  • Helm
  • Kustomize
  • Operators

• Ecosystem / Operations
  • CI/CD
    • GitHub Actions
    • GitLab CI
    • ArgoCD

• Container Runtime
  • containerd
  • CRI-O

• Service Mesh
  • Istio
  • Linkerd

• Cloud Providers
  • Amazon EKS
  • Google GKE
  • Azure AKS
  • DigitalOcean DOKS

Kubernetes = “automated container manager”

—

Imagine you have a web application:

Without Kubernetes:

• You run Docker containers manually
• If one crashes → you restart it yourself
• If traffic increases → you manually add more containers

With Kubernetes:

• You define: “I want 3 instances running”
• Kubernetes ensures 3 instances are always running
• If one crashes → it automatically replaces it
• If traffic increases → it can add more instances automatically

—

Kubernetes is like a restaurant manager:

• You (developer) define the menu (desired state)
• Kubernetes ensures chefs (containers) are always available
• If a chef leaves → it hires a new one automatically