DevOps Classroom notes 10/Jul/2025

Types of containers in a Pod

• In Pod we have three types of containers
  • containers: This is where we run our applications and ideally they would be running all the time. Containers in this will all start in parallel
  • init containers: These are containers which will be created before the containers only after they finish execution the main containers. They execute in sequence. init containers are supposed to check for pre-reqs Refer Here for official docs
• Consider the following spec

---
apiVersion: v1
kind: Pod
metadata:
  name: eight
spec:
  containers:
    - name: web
      image: nginx
      resources:
        requests:
          cpu: 100m
          memory: 10M
        limits:
          cpu: 500m
          memory: 128M
  initContainers:
    - name: cont1
      image: alpine
      args:
        - sleep
        - 5s
    - name: cont2
      image: alpine
      args:
        - sleep
        - 10s

• Now lets create and watch for pod
  
• Refer Here for ephemeral containers

Pod lifecycle

• Refer Here for official docs on pod lifecycle
• Pod Phase Refer Here
  

Controllers

• Pod desired state is to ensure container is running
• Controllers manage Pods Refer Here

ReplicaSet

• Refer Here for official docs.
• Replicaset can maintain replicas of Pods.

Example 1: Create a repicaset with 3 nginx containers

• Refer Here for the changes done

---
apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: nginx
  labels:
    app: nginx
    me: controller
    type: rs
spec:
  minReadySeconds: 5
  replicas: 3
  selector:
    matchLabels:
      app: web
      env: dev
  template:
    metadata:
      labels:
        app: web
        env: dev
        srv: nginx
    spec:
      containers:
        - name: web
          image: nginx

• Now to increase number of replicas we can use kubectl scale, but it is recommended to change yaml and reexecute kubectl apply -f <filepath>