Skip to main content

Advanced Scaling

Enterprise-grade scaling configurations and automation strategies.

Auto-Discovery Configuration​

Cluster Autoscaler Integration​

Enable automatic NodeGroup discovery using tag-based configuration:

--node-group-auto-discovery=vcloud:autoscaler.k8s.io.infra.vnetwork.dev/enabled=true

Configuration Components:

  • Provider: vcloud - Specifies vCloud infrastructure provider
  • Tag Key: autoscaler.k8s.io.infra.vnetwork.dev/enabled
  • Tag Value: true - Enables auto-discovery for tagged NodeGroups

Kubernetes Labels Configuration Kubernetes labels showing autoscaler configuration with automatically managed label

Benefits:

  • Automatic NodeGroup detection by cluster autoscaler
  • Eliminates manual NodeGroup configuration
  • Dynamic scaling based on resource demand
  • Simplified autoscaler setup and maintenance

Note: The autoscaler label autoscaler.k8s.io.infra.vnetwork.dev/enabled is automatically managed:

  • Set to true when autoscaling is enabled
  • Set to false when manual scaling is enabled

Multi-Pattern Discovery​

# Environment-based discovery
--node-group-auto-discovery=vcloud:autoscaler.k8s.io.infra.vnetwork.dev/environment=production

# Workload-specific discovery
--node-group-auto-discovery=vcloud:autoscaler.k8s.io.infra.vnetwork.dev/workload-type=compute-optimized

Multi-Dimensional Scaling​

Horizontal Pod Autoscaler (HPA) Integration​

Combined Scaling Approach:

  • HPA manages replica count based on metrics
  • Cluster autoscaler manages node count based on pod scheduling needs
  • Coordinated operation for optimal resource utilization

HPA Configuration Example:

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: application-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: application-deployment
  minReplicas: 3
  maxReplicas: 100
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 70
  - type: Resource
    resource:
      name: memory
      target:
        type: Utilization
        averageUtilization: 80

Vertical Pod Autoscaler (VPA)​

Resource Right-Sizing:

  • Automatic resource recommendations based on usage patterns
  • Optimal CPU and memory requests/limits
  • Better resource utilization during scaling
  • Cost optimization through accurate allocation

Custom Metrics and Triggers​

Application-Specific Scaling​

Business Metrics:

  • Queue depth for message processing applications
  • Response time for latency-sensitive services
  • Request rate for traffic-based scaling
  • Error rate indicators for resource stress

External Metrics:

  • Database connection pool utilization
  • Cache hit ratio performance
  • Third-party API response times
  • Business event-driven scaling

Predictive Scaling​

Machine Learning-Based Scaling:

  • Historical pattern analysis for recurring usage patterns
  • Seasonal adjustments for daily, weekly, monthly cycles
  • Trend analysis for long-term capacity planning
  • Anomaly detection for unusual patterns

Advanced Verification​

Multi-Layer Verification​

Infrastructure Verification:

# Verify NodeGroup configuration
kubectl get nodegroup NODEGROUP_NAME -o yaml

# Check node labels and taints
kubectl get nodes --show-labels | grep NODEGROUP_SELECTOR

# Verify resource allocation
kubectl describe nodes | grep -A10 "Allocated resources"

# Check cluster autoscaler status
kubectl get configmap cluster-autoscaler-status -n kube-system -o yaml

Application Health Verification:

# Check pod distribution across nodes
kubectl get pods -o custom-columns=NAME:.metadata.name,NODE:.spec.nodeName,STATUS:.status.phase

# Verify service endpoints
kubectl get endpoints SERVICE_NAME -o yaml

# Check application metrics
kubectl top pods --sort-by=cpu

# Validate load balancer configuration
kubectl get services -o wide