K8S

• Components
  • Master node(Control plane)
    • API Server
    • Scheduler
    • Controller Manager
    • etcd
  • Worker node
    • Kubelet
    • kube-proxy
    • Container runtime
• Claster objects
  • Node
    • Pod Scheduling
      • Taints and Tolerations
      • NodeSelector
      • Affinity and Anti-Affinity
  • Pod
    • Init-container
    • Pod Topology Spread Constraints
    • Pod Disruption Budget
  • Service
    • ClusterIP
    • NodePort
    • LoadBalancer
    • ExternalName
  • Deployment
    • Statuses
      • Progressing
      • Replica Failure
      • Complete
  • ConfigMap
  • Secret
  • StatefulSet
  • DaemonSet
  • Ingress
  • Job and CronJob
  • Namespace
    • default
    • kube-node-lease
    • kube-public
    • kube-system
  • Autoscalers
    • Horizontal Pod Autoscaler
    • Vertical Pod Autoscaler
    • Cluster Autoscaler
  • ResourceQuota
  • NetworkPolicy
  • RBAC
    • Role
    • RoleBinding
    • ClusterRole
    • ClusterRoleBinding

• Werf
• Helm

It is recommended not to set CPU limits to avoid performance problems.