docs: add production checklist and rollback plan for my.votcloud.com

infra/deploy/PRODUCTION_CHECKLIST_my.votcloud.com.md

@@ -0,0 +1,80 @@
+# ProxPanel Production Checklist (`my.votcloud.com`)
+
+Date baseline: April 18, 2026  
+Target host: `102.69.243.167`  
+Domain: `my.votcloud.com`
+
+## 1) Pre-Deploy Safety Gate
+
+1. Confirm DNS `A` record points to `102.69.243.167`.
+2. Confirm server ports `80` and `443` are reachable externally.
+3. Confirm current stack health:
+   - `docker compose --env-file /opt/proxpanel/.env.production -f /opt/proxpanel/infra/deploy/docker-compose.production.yml ps`
+   - `curl -fsS http://127.0.0.1:8080/api/health`
+
+## 2) Backup + Rollback Guardrails
+
+1. Create timestamped release directory:
+   - `/opt/proxpanel-backups/<timestamp>/`
+2. Backup application working tree:
+   - `rsync -a /opt/proxpanel/ /opt/proxpanel-backups/<timestamp>/app/`
+3. Backup runtime env:
+   - `cp /opt/proxpanel/.env.production /opt/proxpanel-backups/<timestamp>/env.production.bak`
+4. Backup Postgres before migration:
+   - `docker exec proxpanel-postgres pg_dump -U proxpanel -d proxpanel > /opt/proxpanel-backups/<timestamp>/db_pre_migration.sql`
+
+Rollback entrypoint:
+1. Restore app files from `/opt/proxpanel-backups/<timestamp>/app/`.
+2. Restore env file from `/opt/proxpanel-backups/<timestamp>/env.production.bak`.
+3. Recreate containers with prior source:
+   - `docker compose --env-file /opt/proxpanel/.env.production -f /opt/proxpanel/infra/deploy/docker-compose.production.yml up -d --build`
+4. Restore DB if required:
+   - `cat /opt/proxpanel-backups/<timestamp>/db_pre_migration.sql | docker exec -i proxpanel-postgres psql -U proxpanel -d proxpanel`
+
+## 3) Deploy Order (Strict)
+
+1. Upload release bundle to server.
+2. Extract bundle to `/opt/proxpanel` (keep `.env.production`).
+3. Build and start containers:
+   - `docker compose --env-file .env.production -f infra/deploy/docker-compose.production.yml up -d --build`
+4. Wait for DB + backend readiness.
+5. Apply schema in this order:
+   - `npm run prisma:deploy`
+   - If no migration path or deploy failure, fallback once to `npm run prisma:push`
+6. Run seed:
+   - `npm run prisma:seed`
+7. Health checks:
+   - API health endpoint
+   - Auth login flow
+   - Key pages load
+8. Validate Proxmox sync endpoint with admin token.
+
+## 4) HTTPS + Edge Routing
+
+1. Nginx site config should proxy:
+   - `my.votcloud.com` -> `http://127.0.0.1:<FRONTEND_PORT>`
+2. Issue cert:
+   - `certbot --nginx -d my.votcloud.com --non-interactive --agree-tos -m <ops-email> --redirect`
+3. Verify:
+   - `curl -I https://my.votcloud.com`
+   - Certificate auto-renew timer active.
+
+## 5) Post-Deploy Validation
+
+1. `GET /api/health` returns `200`.
+2. Admin login works.
+3. RBAC page loads users/roles without crash.
+4. Audit logs page renders with no blank state error.
+5. Profile page can list sessions.
+6. System Management page can load branding/policy/CMS data.
+7. Proxmox sync returns success (or actionable credential/SSL error message).
+
+## 6) Incident Rollback Criteria
+
+Rollback immediately if any of the following persists > 10 minutes:
+
+1. API health not green.
+2. Migration errors with broken runtime schema.
+3. Login failures for seeded admin.
+4. Persistent 5xx from `/api/proxmox/sync`.
+5. TLS provisioning failure with site unreachable.