Basic Notes on Containerization

Few Tips :

1. Whenever config-map is updated, you need to manually restart the services to update the changes

2. Ephemeral Volumes exist for the lifespan of a Pod. Persistent Volumes preserve data beyond the lifespan of a Pod.

3. Volume Types

   • emptyDir - Empty directory in Pod with read/write access. Only persisted for the lifespan of a Pod. A good choice for cache implementations or data exchange between containers of a Pod.
   • hostPath: File or directory from the host node’s filesystem. Supported only on single-node clusters and not meant for production.
   • configMap , Secrets: Provides a way to inject configuration data.
   • nfs: An existing NFS (Network File System) share. Preserves data after Pod restart.
   • persistentVolumeClaim : Claims a Persistent Volume.

4. Create a multi-container (easy way )

     kubectl run multipod --image=nginx --dry-run=client -o yaml>multipod.yaml
     kubectl run beta --image=busybox -- command sleep 4800 --dry-run=client -o yaml>>multipod.yaml
   

   vi multipod.yaml

   • Podname: multipod (containers; alpha & beta)
   • update first container name:line 11 –> name: alpha
   • delete below 13-lines (from line-12 onwards)
   • you are good now to apply the yaml file.
   • kubectl describe po multipod

5. Debug the distroless container

      # debian is container image ,fyi 
      kubectl debug <pod-name> exec -it --image=busybox --target=debian        
   

6. Revisit some docker commands

    # Docker build image
   docker build -t  image-name:version . 
   
    # runs container in background and does port forward to 8080      
   docker run -d -p 8080:8080  image-name:version
   
      # login to container shell     
   docker exec -it <container-id> bash
   
   docker tag image-name:version  username/name
   
   docker save -o <file-name.extensions>  image-name:version
   
   docker load --input <file-name.extension>
   

7. Init containers provide initialization logic concerns to be run before the main application even starts.

8. The init container sets up a configuration file in the directory /usr/shared/app. This directory has been shared through a Volume so that it can be referenced by a Node.js-based application running in the main container.

    apiVersion: v1
    kind: Pod
    metadata:
      name: web-app
    spec:
      initContainers:
      - name: configurer
        image: busybox:1.36.1
        command: ['sh', '-c', 'echo Configuring application... && \
                  mkdir -p /usr/shared/app && echo -e "{\"dbConfig\": \
                  {\"host\":\"localhost\",\"port\":5432,\"dbName\":\"customers\"}}" \
                  > /usr/shared/app/config.json']
        volumeMounts:
        - name: configdir
          mountPath: "/usr/shared/app"
      containers:
      - image: iamlaksh1/web-app:1.0.0
        name: web
        ports:
        - containerPort: 8080
        volumeMounts:
        - name: configdir
          mountPath: "/usr/shared/app"
      volumes:
      - name: configdir
        emptyDir: {}

9. Create labels on POD

    kubectl run labeled-pod --image=nginx:1.25.1  --labels=tier=backend,env=dev
    kubectl get pods -l 'team in (shiny, legacy)' --show-labels   #searching 

10. Label selection as part of network-selection API

    apiVersion: networking.k8s.io/v1
    kind: NetworkPolicy
    metadata:
      name: frontend-network-policy
    spec:
      podSelector:
        matchLabels:
          tier: frontend

11. The mapping between the Deployment and the replicas it controls happens through label selection.For label selection to work properly, the assignment of spec.selector.matchLabels and spec.template.metadata needs to match

12. Health Check Attributes

12. Network policies can be aggregated. A default deny rule may disallow ingress and/or egress traffic. An additional network policy can open up those rules with a more fine-grained definition.

13. Service Port Mapping