Pular para conteúdo

Inventário VPS Dev + VPS Master — 2026-08

Snapshot do que roda em cada VPS pra basear os scripts de backup da F4 (DEV-1731). Coletado via SSH em 2026-08-08.

VPS Dev (2.24.117.172)

Disco: /dev/sda1 96 GB, 38 GB usado (39%).

Containers Docker

Container Imagem Papel
cadencia-agenda-app easyappointments:latest EasyAppointments (agenda)
cadencia-agenda-mysql mysql:8.0 DB EasyAppointments
lara-api lara-lara-api Lara nova (Cadencia) — API
lara-worker lara-lara-worker Lara nova — worker
lara-redis redis:7-alpine Lara nova — cache
lara-postgres pgvector/pgvector:pg16 Lara nova — DB (vetorial)

Volumes nomeados relevantes

  • lara-postgres-data — DB da Lara nova (Cadencia)
  • lara-redis-data
  • cadencia-agenda-mysql-data

Crons (user felipe)

  • 0 23 * * * — sync dev-docs
  • */10 * * * * — health-dev selfreport
  • 5 8-22 * * * — pos-briefing trigger

Crons (root)

  • 0 11 * * 1-5 — daily-brief morning
  • 30 21 * * 1-5 — daily-brief evening

Systemd services notáveis

cloudflared, pd-syslog, monarx-agent, docker, ssh.

VPS Master (72.60.4.71)

Disco: /dev/sda1 96 GB, 34 GB usado (35%).

Coolify stack (core)

Container Imagem Papel
coolify coollabsio/coolify:4.1.2 UI + orchestrator
coolify-db postgres:15-alpine Banco interno Coolify (metadata: apps, envs, servers)
coolify-redis redis:7-alpine Cache Coolify
coolify-realtime coolify-realtime:1.0.16 WebSockets UI
coolify-sentinel sentinel:0.0.21 Metrics
coolify-proxy traefik:v3.3.5 Reverse proxy 80/443/8080
nset1suf...-helper coolify-helper:1.0.14 Deploy helper

Apps deployados via Coolify (identificados)

Serviço Container/Volume-chave Observação
Evolution API (WhatsApp Evo) evolution-otklj* + postgres-otklj* + otklj844u...-lara-evolution-db-data Instância principal WhatsApp
Lara legada (GCI) u5n5f0zt...-lara-postgres-data, -lara-redis-data NÃO é a Lara nova — é a legada do grupo GCI
Ecuro middleware ecuromiddleware_postgres_data, yoe8tkelor...-ecuro-middleware-postgres-data Integração cliente GCI-GO
Confirmation-queue (Temporal) kc93gui9...-confirmation-queue-postgres/redis/temporal-postgres-data Sistema fila confirmação agenda
EasyAppointments (Master) cadencia-agenda-db-data, cadencia-agenda-storage Cópia/versão prod da agenda
Obsidian web (?) obsidian_caddy_data, obsidian_caddy_config Caddy servindo algum vault via web
Portainer portainer_data UI Docker adicional
Redis genérico redis-data-j12rhf42* Container j12rhf42* (identificar app depois)

Crons (user master)

  • */5 * * * * — monitor-vps
  • 0 6 * * * — state-aggregator (pd-framework)
  • 30 21 * * * — health-check digest + notify
  • 0 * * * * — health-check apply
  • */30 * * * * — health-check deadman
  • 50 21 * * * — digest heartbeat
  • 15 21 * * * — conferencia-cobranca-pasta worker
  • */10 * * * * — deploy_watcher (Supabase + Slack)
  • 30 7 * * * — self_test_suite (framework)
  • 0 8 * * * — hostinger_watcher (billing/estado VM)
  • 0 9 * * * — confirmacao-agenda worker

Crons (root — cadencia)

  • 0 18 * * 5 — growth_pipeline newsletter (sexta 18h)
  • 0 14 * * * — growth_pipeline sync blog seinfeld instagram
  • 55 13 * * * — retry_provisioning
  • @reboot — mission_control
  • * * * * * — collect-custom-metrics
  • */5 * * * * — cleanup_orphan_ideas
  • 0 12 * * * — drift_check
  • @reboot — resend_webhook
  • */5 * * * * — cadence_tick
  • 0 12 * * * — supabase-advisors
  • 0 12 * * * — check_onboarding_stalled
  • 0 12 * * 1 — audit_resend_domains (segunda)
  • */5 * * * * — evo-pg-guard
  • 0 16 * * * — blog_healthcheck
  • 0 * * * * — reap_stuck_jobs

⚠️ Segurança pendente (não bloqueia backup): VERCEL_TOKEN em texto claro no início do crontab root — Felipe ciente, rotação em separado.

Systemd services notáveis

cadencia-trigger, grafana-webhook, onboarding-consumer (fila webhook→consolidador), alloy (OpenTelemetry), cloudflared, docker, ssh.

Estratégia de backup por fonte (base pra F4)

Fonte Tipo Estratégia
VPS Dev /home/felipe, /home/luiz, /opt filesystem restic via SSH
VPS Dev — DB Lara nova (lara-postgres) container Postgres docker exec lara-postgres pg_dump ... → arquivo → restic
VPS Dev — MySQL EasyAppointments container MySQL docker exec cadencia-agenda-mysql mysqldump ... → arquivo → restic
VPS Dev — Redis Lara container Redis docker exec lara-redis redis-cli SAVE + copiar RDB (dado é cache — vale se dá pra reconstruir?)
VPS Dev — crontabs export crontab -l + sudo crontab -l → arquivo → restic
VPS Master /opt, /etc, /root/scripts filesystem restic via SSH
VPS Master — Coolify metadata banco interno Backup nativo Coolify (dashboard → destination S3-compat na CAIXA1 ou dump manual do coolify-db)
VPS Master — DBs dos apps gerenciados (Evolution, Lara legada, Ecuro, Confirmation-queue, EasyAppointments) container Postgres/MySQL docker exec <db> pg_dump/mysqldump por container identificado
VPS Master — Docker volumes não-DB (storage, config) volumes restic /var/lib/docker/volumes/ (arquivos raw — funciona pra config/storage, não pra DB "quente")
VPS Master — crontabs (user + root) export idem VPS Dev

Perguntas em aberto (não bloqueiam F0)

  1. redis-data-j12rhf42* → qual app dono? (identificar em F4 quando escrever script)
  2. obsidian_caddy_data → algum vault Obsidian exposto via Caddy web? Vale backup?
  3. Confirmation-queue Temporal → Temporal tem workflows em execução — precisa graceful drain antes de dump?

Referências

  • Issue: DEV-1727
  • Issue-mãe: DEV-1726
  • Coleta bruta: ssh vps-dev + ssh -i ~/.ssh/hostinger_prod_master master@72.60.4.71 em 2026-08-08 22:20