Quickstart

Five minutes, three steps. You need a backpulse account (sign up at the portal) and a machine that runs something on a schedule.

1. Install backupctl

backupctl is a single static binary with zero non-stdlib dependencies:

go build -o backupctl ./cmd/backupctl
sudo install -m 0755 backupctl /usr/local/bin/backupctl

2. Create a check

Sign in to the dashboard, create a check, and copy its ping key. The dashboard shows the exact ping URL — it looks like https://backpulse.aniicrite.dev/ping/<key>.

export BACKPULSE_URL=https://backpulse.aniicrite.dev
export BACKPULSE_PING_KEY='<check-key>'

3. Wrap one job

Prefix any scheduled command. backupctl reports start, runs the command, verifies the result, then reports success or failure:

backupctl run --max-age 26h -- restic backup /srv/data

If the backup fails verification — no fresh snapshot, empty artifact — the run exits non-zero and backpulse alerts. A monitoring outage never fails your backup: ping failures only warn.

What next

• Backing up with restic or Borg? Read Monitor restic and Borg backups.
• Running on Kubernetes? Read Monitor Kubernetes CronJobs.
• Dumping Postgres? Read Nightly pg_dump.
• Want the mental model? Read How it works.