Completek8s Classroomnotes 09/Aug/2023

Overview of K8s Workloads

• In Kubernetes we have objects which maintain Pods as part of their desired state
  • Job
  • CronJob
  • DaemonSet
  • ReplicaSet
  • ReplicationController
  • Deployment
  • StatefulSet

Why ReplicaSet over ReplicationController?

• ReplicaSet is successor of Replication Controller and introducted some improvements
  • More Expressive Selectors: In ReplicaSet both Set and Equalty Based Selectors can be used
  • Rolling Updates and Rollbacks: ReplicaSets offer better support
  • Scaling Behavior: RS offer more control over scaling behavior
  • Detailed Status Conditions

Job

• This is a resource that manages execution of specific task or batch job. It ensure that the job completes successfully before considering it done.
• Job might create one or more pods to perform task
• Refer Here
  

CronJob

• This is to run a particular task on schedule
  

Deployment

• This resource manages the deployment and scaling of set of identical pods
• key features
  • Replica Management
  • Rolling Updates
  • Roll backs
  • History and Rollout status
• Deployment internally create replica set which inturn creates pods
  
• Sample

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
  labels:
    app: nginx
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.14.2
        ports:
        - containerPort: 80

StatefulSets

• This resources is designed to manage deployment and scaling of stateful applications such as databases, distrbuted data stores.
• Unlike deployment which is designed for staleless applications, stateful set which is designed for Stateful applications has following features
  • Stable Network Identity
  • Ordered Deployment and Scaling
  • Persistent Storage
  • Headless Services
    

DaemonSet

• This resource ensures that copy of a specific pod runs on each or selected nodes with in cluster
• key features
  • one pod per node
  • node specific tasks
• Use cases:
  • Deploying agents
    • monitoring
    • logging
  • network proxies

Exercise

• Install kubernetes using kube-spray Refer Here
• Try installing a highly available k8s cluster using kubeadm Refer Here