Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/Miguel-Rodriguez15/msvc/llms.txt

Use this file to discover all available pages before exploring further.

The platform uses Kubernetes Horizontal Pod Autoscaler (HPA) for msvc-usuarios and msvc-cursos to respond automatically to CPU demand, and an NGINX Ingress with rate limiting to protect both services from traffic spikes. Together they ensure the system scales out gracefully under load while still rejecting abusive request patterns at the edge.

​
Horizontal Pod Autoscaler

​
HPA Configuration

Both HPA resources are defined using autoscaling/v2 and target CPU utilisation:
Settingmsvc-usuariosmsvc-cursos
API versionautoscaling/v2autoscaling/v2
Minimum replicas11
Maximum replicas55
CPU target (avg utilisation)50%50%
With a CPU request of 400m per pod, the HPA begins scaling out when a pod’s average CPU consumption reaches approximately 200m. It scales back in when utilisation drops back below the threshold, subject to the default stabilisation window.

​
HPA Manifest

The two HPA resources are identical in structure — only the name and scaleTargetRef.name differ: 
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: hpa-msvc-usuarios
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: msvc-usuarios
  minReplicas: 1
  maxReplicas: 5
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 50
For msvc-cursos, the manifest is the same with name: hpa-msvc-cursos and scaleTargetRef.name: msvc-cursos.

​
Monitoring Autoscaling

# See current replica count and CPU targets for all HPAs
kubectl get hpa

# Detailed description including scale events
kubectl describe hpa hpa-msvc-usuarios

# Watch HPA activity in real time
kubectl get hpa -w
The TARGETS column in kubectl get hpa shows <current>/<target> CPU utilisation. It will display <unknown> until the Metrics Server has collected at least one scrape interval of data.
HPA requires the Metrics Server to be running in the cluster. On Minikube, enable it with:
minikube addons enable metrics-server
On a managed cluster (GKE, EKS, AKS), the Metrics Server is typically pre-installed. Verify with kubectl top pods.

​
NGINX Ingress

​
Ingress Rules

All external HTTP traffic enters the cluster at the hostname microservicios.local. The Ingress routes requests to the correct backend service based on the path prefix:
PathBackend ServicePort
/usuariosmsvc-usuarios8001
/cursosmsvc-cursos8002
The rewrite-target: / annotation strips the path prefix before forwarding to the backend, so a request to http://microservicios.local/usuarios/1 reaches msvc-usuarios as GET /1.

​
Rate Limiting Annotations

The Ingress applies per-IP rate limiting to both routes using the following annotations from ingress.yaml:
AnnotationValueDescription
nginx.ingress.kubernetes.io/limit-rps10Maximum 10 requests per second per source IP
nginx.ingress.kubernetes.io/limit-connections5Maximum 5 concurrent connections per source IP
nginx.ingress.kubernetes.io/limit-req-status-code429HTTP status code returned when the rate limit is exceeded
nginx.ingress.kubernetes.io/limit-burst-multiplier3Burst factor — allows up to 30 RPS momentarily before throttling kicks in
nginx.ingress.kubernetes.io/rewrite-target/Strips the path prefix (/usuarios, /cursos) before forwarding to the backend

​
Ingress Manifest

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: microservicios-ingress
  annotations:
    nginx.ingress.kubernetes.io/limit-rps: "10"
    nginx.ingress.kubernetes.io/limit-connections: "5"
    nginx.ingress.kubernetes.io/limit-req-status-code: "429"
    nginx.ingress.kubernetes.io/limit-burst-multiplier: "3"
    nginx.ingress.kubernetes.io/rewrite-target: /
spec:
  ingressClassName: nginx
  rules:
  - host: microservicios.local
    http:
      paths:
      - path: /usuarios
        pathType: Prefix
        backend:
          service:
            name: msvc-usuarios
            port:
              number: 8001
      - path: /cursos
        pathType: Prefix
        backend:
          service:
            name: msvc-cursos
            port:
              number: 8002

​
Enable NGINX Ingress in Minikube

# Enable the NGINX Ingress addon
minikube addons enable ingress

# Verify the Ingress controller pod is running
kubectl get pods -n ingress-nginx
Wait until the ingress-nginx-controller-* pod is in Running state before applying ingress.yaml. After the Ingress resource is created, add the Minikube node IP to your local /etc/hosts file so the microservicios.local hostname resolves: 
echo "$(minikube ip) microservicios.local" | sudo tee -a /etc/hosts
You can then access both services through the Ingress: 
# Users API via Ingress
curl http://microservicios.local/usuarios

# Courses API via Ingress
curl http://microservicios.local/cursos
When a client exceeds the rate limit of 10 RPS sustained (or 30 RPS burst), the NGINX Ingress immediately returns HTTP 429 Too Many Requests. Implement exponential backoff with jitter in your API client to handle this gracefully — for example, retry after 2^attempt * 100ms + random(0..100ms) on each 429 response.

Build docs developers (and LLMs) love

Get started for freeTalk to us