🚀 Guia de Deploy Rápido - Novos Componentes
Data: 2025-12-06
Versão: 1.0
📋 Pré-requisitos
- Docker Swarm ativo
- Acesso ao banco MySQL (csuite)
- Redis rodando
- Traefik configurado
🎯 Componentes para Deploy
- Feature Store - Armazenamento e cache de features
- Model Registry - Versionamento de modelos ML
- ML Pipeline - Pipeline de treinamento (opcional)
- C-Suite Operations - Já deployado (verificar)
🚀 Deploy Automatizado
Opção 1: Script Automático
cd /home/ec2-user/enviroment/apps/c-suite
chmod +x scripts/deploy_new_components.sh
./scripts/deploy_new_components.sh
Opção 2: Deploy Manual
1. Feature Store
cd csuite-feature-store
# Build
docker build -t csuite-feature-store:latest .
# Deploy
docker stack deploy -c docker-stack.yml feature-store
# Verificar
docker service ls | grep feature-store
2. Model Registry
cd csuite-model-registry
# Build
docker build -t csuite-model-registry:latest .
# Deploy
docker stack deploy -c docker-stack.yml model-registry
# Verificar
docker service ls | grep model-registry
3. ML Pipeline (Opcional)
cd scripts/ml_pipeline
# Build
docker build -t ml-pipeline:latest -f Dockerfile .
# Deploy
docker stack deploy -c docker-stack.yml ml-pipeline
# Verificar
docker service ls | grep ml-pipeline
✅ Validação
Script Automático
chmod +x scripts/validate_new_components.sh
./scripts/validate_new_components.sh
Validação Manual
1. Feature Store
# Health check
curl https://csuite.internut.com.br/feature-store/health
# Listar definições
curl https://csuite.internut.com.br/feature-store/definitions
# Sincronizar com Feature Service
curl -X POST https://csuite.internut.com.br/feature-store/sync-definitions
2. Model Registry
# Health check
curl https://csuite.internut.com.br/model-registry/health
# Listar modelos
curl https://csuite.internut.com.br/model-registry/models
3. C-Suite Operations
# Health check
curl https://csuite.internut.com.br/operations/health
# Dashboard
curl https://csuite.internut.com.br/operations/dashboard?org_id=1
🔧 Configuração do Banco de Dados
Feature Store
# Executar schema
mysql -h <host> -u <user> -p csuite < csuite-feature-store/schema.sql
Model Registry
# Executar schema (se houver)
mysql -h <host> -u <user> -p csuite < csuite-model-registry/schema.sql
Agent Coordination & Learning
# Executar schema
mysql -h <host> -u <user> -p csuite < csuite-executive/csuite-api/schema_agent_coordination.sql
📊 Verificação de Status
Listar Serviços
docker service ls | grep -E "feature-store|model-registry|operations"
Ver Logs
# Feature Store
docker service logs feature-store_feature-store-api
# Model Registry
docker service logs model-registry_model-registry-api
# Operations
docker service logs csuite-operations_csuite-operations-api
Verificar Saúde
# Todos os serviços
docker service ps <service_name> --no-trunc
🔗 URLs dos Serviços
| Serviço | URL | Descrição |
|---|---|---|
| Feature Store | https://csuite.internut.com.br/feature-store |
API de features |
| Model Registry | https://csuite.internut.com.br/model-registry |
API de modelos |
| Operations | https://csuite.internut.com.br/operations |
Dashboard operacional |
| Policies UI | https://csuite.internut.com.br/executive/policies/policies.html |
Interface de políticas |
🐛 Troubleshooting
Serviço não inicia
# Ver logs detalhados
docker service logs <service_name> --tail 100
# Verificar variáveis de ambiente
docker service inspect <service_name> --pretty | grep -A 20 Env
Erro de conexão com banco
# Verificar se banco está acessível
mysql -h <host> -u <user> -p -e "SELECT 1"
# Verificar variável CSUITE_DB_HOST no serviço
docker service inspect <service_name> --pretty | grep CSUITE_DB_HOST
Erro 404 no Traefik
# Verificar labels do Traefik
docker service inspect <service_name> --pretty | grep traefik
# Verificar rede
docker network ls | grep superbot-swarm-network
📝 Próximos Passos Após Deploy
- ✅ Validar todos os endpoints
- ✅ Sincronizar Feature Store com Feature Service
- ✅ Testar APIs de coordenação e aprendizado
- ✅ Acessar interface de políticas
- ✅ Configurar monitoramento (se necessário)
🔄 Atualização
Para atualizar um serviço:
# 1. Rebuild
docker build -t <image>:latest .
# 2. Update service
docker service update --image <image>:latest <service_name>
# 3. Verificar
docker service ps <service_name>
Última atualização: 2025-12-06