Endpoints de health check, liveness e readiness probes para monitoramento e orquestracao

Health Check

Endpoints publicos para monitoramento do servico. Nao requerem autenticacao.

Ideais para configurar probes do Kubernetes, Docker health checks e balanceadores de carga.

GET /health

Health check completo com detalhes do servico e status das dependencias (PostgreSQL, Redis, Meta Cloud API).

Auth: Publico.

Curl:

curl http://localhost:8200/health

Resposta (200 OK) — Servico saudavel:

{
  "service": "waba-connector",
  "version": "2.0.0",
  "status": "healthy",
  "dependencies": {
    "database": "healthy",
    "redis": "healthy",
    "meta_api": "healthy"
  }
}

Resposta (200 OK) — Servico degradado:

{
  "service": "waba-connector",
  "version": "2.0.0",
  "status": "degraded",
  "dependencies": {
    "database": "healthy",
    "redis": "unhealthy",
    "meta_api": "healthy"
  }
}

O status sera degraded quando alguma dependencia estiver indisponivel. Operacoes que dependem do componente afetado falharao, mas o servico continua respondendo.

GET /health/live

Liveness probe — verifica se o processo esta vivo e respondendo.

Nao faz verificacoes externas, retornando imediatamente. Usado pelo Kubernetes para reiniciar o container caso ele trave.

Auth: Publico.

Curl:

curl http://localhost:8200/health/live

Resposta (200 OK):

{
  "status": "ok"
}

GET /health/ready

Readiness probe — verifica se o servico esta pronto para receber trafego.

Valida conexao com PostgreSQL, Redis e disponibilidade da Meta Cloud API. Enquanto retornar not_ready, o orquestrador nao deve enviar trafego para esta instancia.

Auth: Publico.

Curl:

curl http://localhost:8200/health/ready

Resposta (200 OK) — Pronto:

{
  "status": "ready",
  "database": "healthy",
  "redis": "healthy",
  "meta_api": "healthy"
}

Resposta (200 OK) — Nao pronto:

{
  "status": "not_ready",
  "database": "healthy",
  "redis": "unhealthy",
  "meta_api": "healthy"
}

Configuracao Kubernetes

Exemplo de Deployment

apiVersion: apps/v1
kind: Deployment
metadata:
  name: waba-connector
spec:
  replicas: 2
  selector:
    matchLabels:
      app: waba-connector
  template:
    metadata:
      labels:
        app: waba-connector
    spec:
      containers:
        - name: waba-connector
          image: waba-connector:latest
          ports:
            - containerPort: 8200
          livenessProbe:
            httpGet:
              path: /health/live
              port: 8200
            initialDelaySeconds: 5
            periodSeconds: 10
            timeoutSeconds: 3
            failureThreshold: 3
          readinessProbe:
            httpGet:
              path: /health/ready
              port: 8200
            initialDelaySeconds: 10
            periodSeconds: 15
            timeoutSeconds: 5
            failureThreshold: 2
          startupProbe:
            httpGet:
              path: /health/live
              port: 8200
            initialDelaySeconds: 3
            periodSeconds: 5
            failureThreshold: 10

Docker Compose Health Check

services:
  connector:
    build: .
    ports:
      - "8200:8200"
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8200/health/live"]
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 10s

Diferencas entre os Probes

Probe	Endpoint	Verificacoes	Uso
Liveness	`/health/live`	Nenhuma (retorno imediato)	Reiniciar container travado
Readiness	`/health/ready`	DB + Redis + Meta API	Controlar trafego do LB
Health	`/health`	DB + Redis + Meta API + metadados	Monitoramento e dashboards

Este guia foi útil?

Última atualização: 7 de abril de 2026

Health Check

On this page