Lab-Pod-Eviction.md

To help students practice the concept of Pod Priority and PriorityClasses in Kubernetes, we can create a hands-on lab where students will create and manage PriorityClass objects and assign them to pods. This lab will demonstrate how pod priority affects scheduling and eviction behavior in resource-constrained environments.

Here’s a step-by-step guide for the lab:

---

Lab: Practice with Kubernetes Pod Priority and PriorityClasses

Prerequisites:

• A working Kubernetes cluster
• kubectl CLI tool installed and configured to interact with your cluster.

---

Lab Setup Steps:

1. Create Priority Classes

In this step, students will create different PriorityClass resources to assign priorities to pods.

Task:

• Create three different PriorityClasses with varying priority values:
  • high-priority: A high priority value for critical workloads.
  • medium-priority: A moderate priority value.
  • low-priority: A low priority value for non-essential workloads.

Instructions:

1. Create a YAML file priority-classes.yaml:

apiVersion: scheduling.k8s.io/v1
kind: PriorityClass
metadata:
  name: high-priority
value: 1000000
globalDefault: false
description: "High priority pods for critical workloads"
---
apiVersion: scheduling.k8s.io/v1
kind: PriorityClass
metadata:
  name: medium-priority
value: 500000
globalDefault: false
description: "Medium priority pods"
---
apiVersion: scheduling.k8s.io/v1
kind: PriorityClass
metadata:
  name: low-priority
value: 100000
globalDefault: false
description: "Low priority pods"

2. Apply the PriorityClass resources to the cluster:

kubectl apply -f priority-classes.yaml

3. Verify the PriorityClasses have been created:

kubectl get priorityclass

This should list the three PriorityClasses (high-priority, medium-priority, low-priority).

---

2. Create Pods with Different Priority Classes

Now, students will create pods that use the PriorityClass resources they just created.

Task:

• Create three pods:
  • A high-priority pod.
  • A medium-priority pod.
  • A low-priority pod.

Instructions:

1. Create a YAML file pods-with-priority.yaml:

apiVersion: v1
kind: Pod
metadata:
  name: high-priority-pod
spec:
  priorityClassName: high-priority
  containers:
  - name: nginx
    image: nginx
    resources:
      requests:
        memory: "64Mi"
        cpu: "250m"
      limits:
        memory: "128Mi"
        cpu: "500m"

---
apiVersion: v1
kind: Pod
metadata:
  name: medium-priority-pod
spec:
  priorityClassName: medium-priority
  containers:
  - name: nginx
    image: nginx
    resources:
      requests:
        memory: "64Mi"
        cpu: "250m"
      limits:
        memory: "128Mi"
        cpu: "500m"

---
apiVersion: v1
kind: Pod
metadata:
  name: low-priority-pod
spec:
  priorityClassName: low-priority
  containers:
  - name: nginx
    image: nginx
    resources:
      requests:
        memory: "64Mi"
        cpu: "250m"
      limits:
        memory: "128Mi"
        cpu: "500m"

2. Apply the pod configuration:

kubectl apply -f pods-with-priority.yaml

3. Verify that the pods have been created:

kubectl get pods

The output will show that the pods high-priority-pod, medium-priority-pod, and low-priority-pod are running.

---

3. Simulate Resource Pressure (Optional for Eviction)

Next, students will simulate resource pressure on the cluster by running resource-intensive pods and observing how Kubernetes evicts lower-priority pods.

Task:

• Simulate resource pressure by creating BestEffort pods that consume a lot of resources and trigger the eviction process.

Instructions:

1. Create a YAML file best-effort-pods.yaml for resource-hogging pods:

apiVersion: v1
kind: Pod
metadata:
  name: resource-hogger-pod-1
spec:
  containers:
  - name: stress-test
    image: polinux/stress
    resources:
      requests:
        memory: "1Gi"
        cpu: "500m"
      limits:
        memory: "2Gi"
        cpu: "1"

---
apiVersion: v1
kind: Pod
metadata:
  name: resource-hogger-pod-2
spec:
  containers:
  - name: stress-test
    image: polinux/stress
    resources:
      requests:
        memory: "1Gi"
        cpu: "500m"
      limits:
        memory: "2Gi"
        cpu: "1"

2. Apply the resource-hogging pods to the cluster:

kubectl apply -f best-effort-pods.yaml

3. Monitor the pod eviction process by checking the pod status:

kubectl get pods

• Pods with lower priority, like the low-priority-pod, may be evicted first to make room for higher-priority pods when the node reaches its resource limits.

4. Verify Pod Evictions:

   • Use kubectl describe pod <pod-name> to check the reason for eviction, e.g.,:

   kubectl describe pod low-priority-pod

   The Event section will show eviction details like:

   Evicted: The node was under pressure, and pod was evicted
   

---

4. Clean Up

Once the practice is complete, students should clean up the resources they’ve created to keep the cluster clean.

Task:

• Delete the created pods and PriorityClasses.

Instructions:

1. Delete the created pods:

kubectl delete pod high-priority-pod medium-priority-pod low-priority-pod

2. Delete the resource-hogging pods:

kubectl delete pod resource-hogger-pod-1 resource-hogger-pod-2

3. Delete the PriorityClasses:

kubectl delete priorityclass high-priority medium-priority low-priority

---

Conclusion:

In this lab, students practiced:

• Creating and managing PriorityClass resources.
• Assigning different PriorityClasses to pods.
• Simulating node resource pressure to observe how Kubernetes handles pod evictions.
• Observing how higher-priority pods are less likely to be evicted than lower-priority pods under pressure.