• Context: A group of access parameters to a Kubernetes cluster
  • Cluster
    • Nodes: Physical or virtual machines
      • Master node/control plane
      • Worker nodes
      • Contains:
        1. Container runtime
        2. Kubelet: Manages the pod lifecycle
        3. Kube-proxy: Manages network rules
        4. Pods
  • User
  • Namespace: Group resources (e.g., Dev, Test, Prod)
• Pods: Smallest deployable units
  • Encapsulate containers
  • Can run multiple containers
  • States:
    • Pending
    • Running
    • Succeeded
    • Failed
    • Unknown
    • CrashLoopBackOff
• Labels: Key-value pairs for identifying resources
• Selectors: Filter or select objects using labels
• Services: Provide stable IP addresses
• Containers: Share IP and storage within a pod
• Workloads: Application running on Kubernetes
  • Pod
  • ReplicaSet: Manages pod replicas
  • Deployment: Manages pod templates
  • StatefulSet: For stateful applications
  • DaemonSet: Ensures pods on all nodes
  • Job: Short-lived tasks
  • CronJob: Scheduled tasks
• Deployments
  • Replicas: Number of pod instances
  • Strategy
    • RollingUpdate
    • Recreate
• StatefulSet: For stateful applications (e.g., databases)
• Services
  • ClusterIP: Internal visibility
  • NodePort: External visibility
  • LoadBalancer: External load balancing
  • Ingress: Advanced routing
• Volumes
  • Persistent Volume
  • Persistent Volume Claim
  • Reclaim Policies: Delete, Retain
• Access Modes
  • ReadOnlyMany
  • ReadWriteMany
  • ReadWriteOnce
• Persistent Volume States
  • Available
  • Bound
  • Released
  • Failed
• ConfigMaps: Externalize configuration
• Horizontal Pod Autoscaling
  • Requires metrics server
  • Scale based on metrics

kubectl Commands

• config
  • current-context: Get current context
  • get-contexts: List all contexts
  • use-context context name: Set current context
  • set-context --current --namespace=namespace name
  • delete-context context name
  • rename-context old-name new-name
• create
  • deployment name --image=image name: Imperative way
  • -f yaml file: Declarative way
  • ns namespace name
• delete
  • all --all: Delete all pods
  • deployment name
  • ns namespace name
  • pod podname
  • -f pod-definition.yaml
  • rs rs name
  • ds ds name
  • sts sts name
  • job job name
  • cj job name
  • svc service name
  • pv pv name
  • pvc pvc name
  • hpa hpa name
• get
  • namespaces or ns: List namespaces
  • pods -o wide
  • nodes
  • ep service name: Service endpoint IP
  • rs: List ReplicaSets
  • ds: List DaemonSets
  • sts: List StatefulSets
  • pvc: Persistent volume claims
  • job
  • node node name
  • pod pod name
  • rs rs name
  • ds ds name
  • cj
  • svc -o wide
  • nodes -o wide: Node IP addresses
  • pv
  • pvc
  • cm: ConfigMap
  • hpa hpa name
• exec
  • -it podname -- sh: Interactive shell
• run podname --image=image name
• logs podname
• rollout
  • status: Update progress
  • history deployment name: Deployment history
  • undo name: Rollback deployment
• describe
  • svc service name
  • pv pv name
  • pvc pvc name
• scale resource-type/resource-name --replicas=number