⚙️ 1. Resource Management Testing
One of the most powerful features in Kubernetes is the ability to control how much CPU and memory a container is allowed to use. Without limits, a single container could consume all the node resources and impact other applications running in the cluster.
| Setting | Purpose |
|---|---|
| Request | Minimum resources Kubernetes guarantees for the container |
| Limit | Maximum resources the container is allowed to consume |
📄 Resource Test Deployment File
File: 08-resource-test.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: resource-test
spec:
replicas: 1
selector:
matchLabels:
app: resource-test
template:
metadata:
labels:
app: resource-test
spec:
containers:
- name: stress-container
# This container intentionally generates CPU load
# It allows us to observe how Kubernetes enforces limits
image: polinux/stress
# Run stress tool to consume CPU
args: ["--cpu", "1"]
resources:
# Guaranteed CPU allocation
requests:
cpu: "250m"
# Maximum CPU allowed
limits:
cpu: "500m"
This deployment launches a container that intentionally generates CPU load using the stress utility. Because we defined both requests and limits, Kubernetes can enforce resource rules.
- 🧠 250m request means the container is guaranteed 0.25 CPU.
- 🚦 500m limit means it cannot exceed half a CPU core.
- ⚡ If it tries to use more than that, Kubernetes will throttle the container.
🚀 Apply the Deployment
kubectl apply -f 08-resource-test.yaml
This command creates the deployment and schedules the pod on a node. The container will immediately begin generating CPU load.
👀 Watch the Pod in Real Time
kubectl get pods -l app=resource-test -w
This command streams pod status updates in real time.
- -l filters pods using a label selector
- -w enables live watch mode
🔎 Inspect Pod Resource Settings
kubectl describe pod -l app=resource-test
This command displays detailed pod information including:
- Configured CPU requests and limits
- Events related to resource throttling
- Container lifecycle events
📊 Check Real-Time Resource Usage
kubectl top pod -l app=resource-test
This shows the actual CPU and memory usage of the running pod.
❤️🩹 2. Health Check Testing
Kubernetes continuously monitors application health using probes. These checks help Kubernetes automatically detect and recover from application failures.
| Probe | Purpose |
|---|---|
| Readiness Probe | Determines if the pod is ready to receive traffic |
| Liveness Probe | Restarts containers if they become unhealthy |
📄 Health Check Deployment File
File: 06-health-checks.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-health
spec:
replicas: 1
selector:
matchLabels:
app: nginx-health
template:
metadata:
labels:
app: nginx-health
spec:
containers:
- name: nginx
image: nginx:1.14
ports:
- containerPort: 80
# Readiness probe checks if pod can receive traffic
readinessProbe:
httpGet:
path: /
port: 80
initialDelaySeconds: 5
periodSeconds: 5
# Liveness probe ensures container is still healthy
livenessProbe:
httpGet:
path: /
port: 80
initialDelaySeconds: 15
periodSeconds: 10
This deployment uses two important health mechanisms:
- 🟢 Readiness probe prevents traffic from reaching a container until it is ready.
- 🔁 Liveness probe automatically restarts the container if it becomes unhealthy.
🚀 Deploy the Health Check Example
kubectl apply -f 06-health-checks.yaml
👀 Watch Pod Status
kubectl get pods -l app=nginx-health -w
Watching pods live allows you to see when:
- Probes succeed
- Probes fail
- Containers restart automatically
🔎 Inspect Probe Configuration
kubectl describe pod -l app=nginx-health
This command shows detailed probe configuration including:
- Probe timing settings
- Failure thresholds
- Probe success/failure events
📜 Check Container Logs
kubectl logs -l app=nginx-health
Because probes send HTTP requests to nginx, these requests appear inside the nginx access logs. This makes debugging health check issues much easier.
🧹 3. Cleanup Commands
After testing is complete, it’s good practice to remove temporary resources from your cluster.
kubectl delete -f 08-resource-test.yaml kubectl delete -f 06-health-checks.yaml
These commands remove both deployments and automatically terminate their associated pods.
✅ Verify Cleanup
kubectl get all
This command lists all resources in the current namespace. After cleanup, you should only see the default kubernetes service.