WAL-G backup staleness monitor #6
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Every hour, check latest WAL-G backup timestamps per DB (immich, lobe) in Garage S3.
If last backup >26h old → Gotify priority 8 alert.
Context: silent WAL-G stall went undetected for 13h in 2026-05-22 incident.
Nodes: Cron → Execute Command (mc ls garage S3 paths) → IF age >26h → Gotify.
Closed by /usr/local/sbin/walg-staleness-check.sh + walg-staleness.timer on CT 103 (hourly). Queries pg_stat_archiver on each Postgres host; alerts on stale archive OR recent failures. Found Immich wal-g had been broken for 42h (garage hostname unresolvable after Garage container was manually recreated on bridge network not shared_backend) — fixed by reconnecting garage to shared_backend.