Sync Strategies e Health Checks

Estratégias de Sincronização

apiVersion: argoproj.io/v1alpha1
kind: Application
spec:
  syncPolicy:
    automated:
      prune: true
      selfHeal: true
      # Outras estratégias:
      # allowEmpty: true  — permite app sem recursos
      # pruneLast: true  — prune após sync
    syncOptions:
      - CreateNamespace=true
      - ApplyOutOfSyncOnly=true
      - PruneLast=true
      - PrunePropagationPolicy=foreground
      - RespectIgnoreDifferences=true

Sync Phases

Sync:
├── PreSync   (hooks: job de migration)
├── Sync      (apply resources)
│   ├── manifests
│   ├── hooks
│   └── ...
└── PostSync  (hooks: smoke tests, notifications)

Sync Hooks

apiVersion: batch/v1
kind: Job
metadata:
  name: db-migration
  annotations:
    argocd.argoproj.io/hook: PreSync
    argocd.argoproj.io/hook-delete-policy: HookSucceeded
spec:
  template:
    spec:
      containers:
        - name: migration
          image: my-app:latest
          command: ["./run-migrations.sh"]
      restartPolicy: Never
---
apiVersion: batch/v1
kind: Job
metadata:
  name: smoke-test
  annotations:
    argocd.argoproj.io/hook: PostSync
    argocd.argoproj.io/hook-delete-policy: HookSucceeded
spec:
  template:
    spec:
      containers:
        - name: test
          image: curlimages/curl
          command: ["curl", "-f", "http://app-service/health"]
      restartPolicy: Never

Hooks disponíveis:
├── PreSync    → antes de aplicar recursos
├── Sync       → durante (substitui recurso)
├── PostSync   → após sync bem-sucedido
├── SyncFail   → quando sync falha
└── Skip       → não executa

Health Checks Customizados

# argocd-cm ConfigMap
apiVersion: v1
kind: ConfigMap
metadata:
  name: argocd-cm
  namespace: argocd
data:
  resource.customizations: |
    networking.k8s.io/Ingress:
      health.lua: |
        hs = {}
        hs.status = "Healthy"
        return hs
    batch/v1/CronJob:
      health.lua: |
        if obj.status ~= nil and obj.status.lastScheduleTime ~= nil then
          hs = {status = "Healthy"}
        else
          hs = {status = "Progressing"}
        end
        return hs

Sync Status

Status do Sync:
├── Synced          → Git = Cluster
├── OutOfSync       → Git ≠ Cluster (drift)
├── Syncing         → sync em andamento
└── Failed          → sync falhou

Saúde:
├── Healthy         → tudo ok
├── Progressing     → deploy rolando
├── Degraded        → algo errado
├── Suspended       → pausado (HPA scaled to 0)
└── Missing         → recurso não encontrado

Selective Sync

# Sincronizar apenas recursos específicos
argocd app sync guestbook --resource apps/Deployment:guestbook
argocd app sync guestbook --label app=web

# Sync por label
argocd app sync -l app.kubernetes.io/instance=guestbook

# Sync com prunção manual
argocd app sync guestbook --prune

Compare With

# Comparar estado atual com Git
argocd app diff guestbook
argocd app diff guestbook --local ./manifests/

# Ver diferenças específicas de recurso
argocd app diff guestbook --resource apps:Deployment:api

Estratégia de sync = automated + prune + selfHeal + ApplyOutOfSyncOnly. Hooks permitem migration e smoke tests no sync cycle. Health checks customizados para recursos não-padrão.