Heartbeat monitoring

Lookout watches your server from the inside, but if the whole machine goes down, the network drops, or the Lookout process dies, the agent can't send you an alert. A heartbeat monitor closes that gap: Lookout expects a regular ping, and alerts you when one doesn't arrive.

This is a dead-man's switch built into Lookout — there's no third-party service to sign up for.

How it works

1. You create a heartbeat monitor in Lookout and get a unique ping URL.
2. The Lookout agent pings that URL on a fixed interval.
3. Lookout records each ping. As long as they keep arriving, the monitor is up.
4. If a ping is late beyond its grace period, the monitor goes down and Lookout alerts you.
5. When pings resume, Lookout sends a recovery alert.

Because Lookout uses the time it receives each ping, clock skew on your server doesn't matter.

1. Create a monitor

Open your Lookout dashboard and click New monitor. Give it:

• a name (e.g. "Production VPS")
• an expected interval — how often the agent pings (default 60s)
• a grace period — how much lateness to allow before alerting (default 120s)

With the defaults, a monitor pinging every minute is marked down after about three minutes of silence.

2. Copy the ping URL

After creating the monitor, Lookout shows its unique ping URL once:

https://lookout.kelvinamoaba.com/ping/lk_ping_xxxxxxxxxxxx

Treat this URL like a secret — anyone who has it can send pings on your behalf. If it ever leaks, rotate it from the monitor's page; the old URL stops working immediately.

3. Add it to your agent config

Set it as the heartbeat.url in /etc/lookout/config.yaml, with the interval matching what you entered when creating the monitor:

# /etc/lookout/config.yaml
heartbeat:
  url: "https://lookout.kelvinamoaba.com/ping/lk_ping_xxxxxxxxxxxx"
  interval: 60s

Then restart Lookout:

sudo systemctl restart lookout

The first ping should arrive within a minute and your monitor will flip to up.

Monitor states

State	Meaning
Pending	Created, but no ping received yet. Waiting for the first ping.
Up	A ping arrived recently enough.
Late	Past the expected interval but still inside the grace period — no alert yet.
Down	Past the grace period. Lookout fires a down alert.
Paused	Alerting is temporarily disabled. Pings are still recorded.

Alerts

A heartbeat monitor sends two kinds of alert:

• Down — when pings stop arriving within the grace window.
• Recovery — when a new ping arrives after the monitor was down.

A brand-new monitor that has never been pinged stays pending and never alerts, so you won't get noise before setup is finished. Paused monitors don't alert either.

Notification channels

Heartbeat alerts are sent from Lookout's servers — your machine is down, so it can't alert you itself. Choose where they go under Settings in the dashboard: Discord, Google Chat, or a generic webhook. Webhook URLs must be public https endpoints.

This is separate from the agent's own notifiers (Discord, Slack, email, and the rest) that deliver its threshold alerts — those run on the server and can't fire once it goes offline, which is why the heartbeat is delivered from Lookout instead.

Tips

• The grace period absorbs a single missed ping, so a brief hiccup won't trip a false alarm.
• Pause a monitor before planned maintenance to avoid a down alert, then resume it afterward.
• Duplicate or very frequent pings are harmless.