- backup.sh: fix openclaw (tar-eaba un hostPath stub vacio, 104 bytes/dia
desde hace meses); ahora resuelve el PVC real via local-path.
- Nuevos servicios: ghost-en/zona-exclusion (VACUUM INTO consistente via
node del pod + tar del content), gitea (sqlite .backup + repos + conf,
sin packages 6.4G reconstruibles), researchowl (sqlite backup API via
python3), roswell-pg, umami-pg, infisical-pg, trilium, vaultwarden,
uptime-kuma (sqlite3 .backup), filebrowser-db, home-master (sin ~/.ssh
ni caches), k3s token + /etc/rancher/k3s (cadena de restauracion).
- Validacion por artefacto: gzip -t + tamano minimo; el job sale 1 si
falta algo -> alerta Grafana->Telegram existente.
- rclone-mega: de semanal a diario (03:30); --exclude reddit-intel/.
- Nuevo CronJob reddit-intel-backup en n97 (unico dato del worker fuera
de k3s; su cron local escribia en el mismo disco).
- verify.sh: lista completa (20 servicios + reddit-intel), diario,
ventana de frescura 3 dias (BusyBox-safe).
- backoffLimit 1 (retry recrearia artefactos y romperia la retencion).
- RESTORE.md: runbook con la cadena token->state.db->infisical-secrets.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Add hostPath volume for /var/lib/rancher/k3s/server/db (readOnly)
- Script copies state.db + WAL files → k3s-db_<date>.tar.gz in /data/backups/backups/
- Rotation: keeps last 7 copies (same policy as other services)
- rclone-mega-backup picks it up automatically (syncs full /data/backups/backups/)
- Also tracks the CronJob manifest in git (was previously untracked)
Note: k3s uses SQLite/kine (not embedded etcd). etcd-snapshot is disabled.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
chemavxandClaude Fable 5