• Well configured containers let Kubernetes know how much CPU and RAM to assign to a worker node.
  • resources refer to the amount of CPU and memory for the worker node running your pods.
  • If you deploy a pod without a set of resource requests, it can be scheduled on a node that does not have enough processing power or memory. The node can fail because of this.
  • You can also start a pod on a node, but the application can use all of the resources on the node, causing it to fail. This causes outages.
  • How to add resource requests and limits to our pod info container.
  • Example resource limitation yaml file: 
      resources:
    requests:
      memory: "64Mi"
      cpu: "250m"
    limits:
      memory: "128Mi"
      cpu: "500m"
  • We can add this to the deployment.yaml file like so: 
    --- 
apiVersion: apps/v1
kind: Deployment
metadata:
name: pod-info-deployment
namespace: development
labels:
  app: pod-info
spec:
replicas: 1
selector:
  matchLabels:
    app: pod-info
template:
  metadata:
    labels:
      app: pod-info
  spec:
    containers:
    - name: pod-info-container
      image: kimschles/pod-info-app:latest
          resources:
                requests:
                    memory: "64Mi"
                    cpu: "250m"
                limits:
                    memory: "128Mi"
                    cpu: "500m"
      ports:
      - containerPort: 3000
      env:
        - name: POD_NAME
          valueFrom:
            fieldRef:
              fieldPath: metadata.name
        - name: POD_NAMESPACE
          valueFrom:
            fieldRef:
              fieldPath: metadata.namespace
        - name: POD_IP
          valueFrom:
            fieldRef:
              fieldPath: status.podIP
  • Now, we have set resource requests and limits for the pod-info container.
  • To explain the resources sequence:
    • The request section says to not schedule the pod, unless the node has at least 64Mi (Mi is mebibyte). A mebibyte is 1,048,576 bytes.
    • It also needs 250m of CPU. m is for milliCPU.
    • The limits section says to stop using this container, if it exceeds 128Mi of memory and 500m of CPU.
  • How do you decide on these values, it depends on the container you are running in the pod.
  • To update the pods, we then run: 
    kubectl apply -f deployment.yaml
  • If you receive an output such as: 
    howard@skwigelf:~$ kubectl get pod -n development
NAME                                   READY   STATUS        RESTARTS      AGE
pod-info-deployment-7587d5cc86-lkccq   1/1     Terminating   1 (25h ago)   4d19h
pod-info-deployment-7587d5cc86-vk6br   1/1     Running       1 (25h ago)   4d19h
pod-info-deployment-7587d5cc86-zs4kp   1/1     Terminating   1 (25h ago)   4d19h
  • The old pod-info deployment pods are terminating.
  • New pod-info deployment pods are running and this is because we ran the deployment.yaml script again.
    • These new pods would have the resource requests and limits.
    • Important step to avoid outages.

Updated: