100-Days-Of-DevOps-Challenge-KodeKloud

Set Resource Limits in Kubernetes Cluster

The Nautilus DevOps team has noticed performance issues in some Kubernetes-hosted applications due to resource constraints. To address this, they plan to set limits on resource utilization. Here are the details:

Create a pod named httpd-pod with a container named httpd-container. Use the httpd image with the latest tag (specify as httpd:latest). Set the following resource limits:
Requests: Memory: 15Mi, CPU: 100m
Limits: Memory: 20Mi, CPU: 100m

Note: The kubectl utility on jump_host is configured to operate with the Kubernetes cluster.

Steps

Let’s create pod.yaml file with with these contents. Our updated file will be look like this:

apiVersion: v1
kind: Pod
metadata:
    name: httpd-pod
    labels:
        app: httpd_pod
spec:
    containers:
        - name: httpd-container
          image: httpd:latest
          ports:
            - containerPort: 80
          resources:
            limits:
                cpu: 100m
                memory: 20Mi
            requests:
                cpu: 100m
                memory: 15Mi

Let’s create the pod:
```
  kubectl apply -f pod.yml
```
Verify pod:
```
  kubectl get pods
```

Good to Know?

Resource Management

Requests: Minimum resources guaranteed to container
Limits: Maximum resources container can use
Quality of Service: BestEffort, Burstable, Guaranteed
Resource Types: CPU (cores), Memory (bytes), Storage

CPU Resources

Units: Cores (1000m = 1 core)
Fractional: 100m = 0.1 core
Throttling: CPU limits enforced by throttling
Scheduling: Requests used for pod placement

Memory Resources

Units: Bytes (Ki, Mi, Gi)
OOM Killer: Exceeding limits triggers container kill
Non-compressible: Memory cannot be throttled
Eviction: Node pressure can evict pods

Resource Planning

Monitoring: Track actual resource usage
Right-sizing: Set appropriate requests and limits
Node Capacity: Consider node resource availability
Cluster Autoscaling: Scale nodes based on resource needs

This site is open source. Improve this page.