Resource Management & Limits¶

🎯 Learning Objectives

Master resource requests and limits
Understand resource quotas and limit ranges
Learn advanced resource management patterns
Troubleshoot resource-related issues
Optimize resource utilization

Effective resource management is critical for cluster stability and performance. Understanding requests, limits, quotas, and limit ranges enables optimal resource utilization.

Resource Requests vs Limits

Requests: Guaranteed resources (used for scheduling) Limits: Maximum resources (enforced by cgroups)

OOMKilled

Containers exceeding memory limits are killed (OOMKilled). Always set appropriate limits.

Resource Requests and Limits¶

Basic Resource Configuration¶

apiVersion: v1
kind: Pod
metadata:
  name: resource-demo
spec:
  containers:
  - name: app
    image: nginx:alpine
    resources:
      requests:
        memory: "64Mi"
        cpu: "250m"
      limits:
        memory: "128Mi"
        cpu: "500m"

CPU Units

CPU can be specified as: whole numbers (1, 2) or millicores (1000m = 1 CPU).

Quality of Service Classes¶

Kubernetes assigns QoS classes based on resource configuration:

Guaranteed: - All containers have requests = limits - Highest priority, last to be evicted

Burstable: - At least one container has request < limit - Medium priority

BestEffort: - No requests or limits - Lowest priority, first to be evicted

QoS Strategy

Use Guaranteed QoS for critical workloads, Burstable for most applications, BestEffort only for non-critical.

Resource Quotas¶

Namespace Quota¶

apiVersion: v1
kind: ResourceQuota
metadata:
  name: compute-quota
  namespace: production
spec:
  hard:
    requests.cpu: "10"
    requests.memory: 20Gi
    limits.cpu: "20"
    limits.memory: 40Gi
    persistentvolumeclaims: "10"
    pods: "20"

Quota Enforcement

Resource quotas are enforced at namespace level. Plan quotas carefully to avoid blocking deployments.

Limit Ranges¶

Default Limits¶

apiVersion: v1
kind: LimitRange
metadata:
  name: mem-limit-range
  namespace: default
spec:
  limits:
  - default:
      memory: "512Mi"
      cpu: "500m"
    defaultRequest:
      memory: "256Mi"
      cpu: "250m"
    type: Container

Default Limits

Limit ranges provide defaults when containers don't specify resources. Prevents resource exhaustion.

Troubleshooting¶

Common Issues¶

Pods Pending (Insufficient Resources)¶

# Check node resources
kubectl describe node <node-name>

# Check resource requests
kubectl get pods -o custom-columns=NAME:.metadata.name,REQ-CPU:.spec.containers[*].resources.requests.cpu,REQ-MEM:.spec.containers[*].resources.requests.memory

# Check quota usage
kubectl describe quota -n <namespace>

Troubleshooting Steps

Check node capacity
Verify resource requests
Check resource quotas
Review pod scheduling events
Consider node scaling

OOMKilled Pods¶

# Check pod status
kubectl get pods | grep OOMKilled

# Check memory usage
kubectl top pod <pod-name>

# Check limits
kubectl describe pod <pod-name> | grep -A 5 Limits

Memory Issues

OOMKilled indicates memory limit exceeded. Increase limits or optimize application memory usage.

Best Practices¶

Production Recommendations

Always set resource requests and limits
Use Guaranteed QoS for critical workloads
Implement resource quotas per namespace
Set limit ranges for defaults
Monitor resource usage
Right-size resources based on metrics

Next Chapter: Performance Tuning