Get a push notification when a backup finishes

Report success, size, and failure

A good backup notification tells you three things: that it ran, how big the result was, and — if it broke — where. A bash trap on ERR handles the failure path, so an error anywhere in the script fires a POST before it exits:

#!/usr/bin/env bash
set -euo pipefail
LERT_URL="https://lert.dev/p/your-id"
OUT="/backups/db-$(date +%F).sql.gz"

notify() {
  curl -sS -X POST "$LERT_URL" \
    -H "Content-Type: application/json" \
    -d "$(jq -n --arg t "$1" --arg b "$2" \
        '{title:$t, body:$b, tag:"backup"}')" >/dev/null
}

# Any error below fires a failure ping with the line number
trap 'notify "Backup FAILED ❌" "exit on line $LINENO"' ERR

pg_dump "$DATABASE_URL" | gzip > "$OUT"

SIZE=$(du -h "$OUT" | cut -f1)
notify "Backup OK ✅" "$OUT — $SIZE in ${SECONDS}s"

set -euo pipefail makes the script stop on the first error and treats a failure anywhere in the pg_dump | gzip pipe as a failure. The trap ... ERR catches that and sends the failure notification with the line number; if everything works, the last line reports the path, human-readable size, and how many seconds it took.

Set it up in three steps

1. Copy your Lert URL

Open the Lert app and copy your unique endpoint — something like https://lert.dev/p/your-id. Store it in the script or an environment variable and treat it like a password.

2. Wrap your existing backup command

Whatever you run now — pg_dump, mysqldump, tar, restic, rsync — goes in the middle. Add set -euo pipefail at the top, the ERR trap for failures, and the size-and-duration POST at the end.

3. Schedule it

Run it from cron overnight. Because the script reports both outcomes, you'll get a small green "Backup OK" every morning — and the absence of one is a signal in itself.

# 2:30am every night
30 2 * * *  /opt/scripts/nightly-backup.sh

Guard against the silent skip

The failure mode a nightly ping catches best is the one no error handler sees: the job that stopped being scheduled at all. A crashed cron, a full disk that killed the daemon, a host that got rebuilt without the crontab. When the expected 2:30am "Backup OK" doesn't arrive, that gap is your cue to go look. Some teams take it further with a dead-man's-switch: a separate check that alerts if no success ping was recorded overnight.

Where this earns its keep

• Database dumps: nightly pg_dump / mysqldump with the compressed size confirmed each morning.
• File & volume backups: restic, borg, or tar snapshots reporting how much was written.
• Offsite sync: an rsync or rclone push to S3 that pings when the upload completes.
• Restore drills: wrap a periodic test-restore so you're told it still actually works.

Why Lert for backup alerts

Backup tooling that emails on failure trains you to trust silence — and silence is exactly what a dead backup produces. Lert flips it: a real push every night that a locked phone can't miss, set up with one function and one POST. No agent on the box, no account, no key to rotate — the endpoint URL is the authentication. It's the cheapest insurance you'll add all year.