# Last Seen Up

> Uptime, heartbeat, SSL certificate and domain expiry monitoring. It watches
> your sites, servers and scheduled jobs around the clock, double-checks every
> failure before it wakes anyone, then sends the alert straight to your phone.

Last Seen Up is a hosted service at https://lastseenup.com. There is no
open-source edition, no self-hosted edition and no downloadable server. Two ways
to use it: a web dashboard, and a native iOS app.

The iOS app is on the App Store:
https://apps.apple.com/us/app/last-seen-up/id6797648903 — free, requires iOS 18
or later, and included with every plan. It signs into the same account as the
dashboard, so create the account either place.

## What it checks

- **HTTP** — loads a page or API and checks the status code, headers and content you expect. Catches a bad deploy, a 500, a response that quietly changed.
- **TCP** — opens a connection to a port. Catches a service that died without taking the server with it.
- **Ping** — sends an ICMP ping to a host. Catches a machine that has dropped off the network.
- **DNS** — looks up records and compares them to what they should be. Catches a DNS change nobody mentioned.
- **SSL** — checks the certificate being served and the days it has left. Catches an expiry weeks before visitors see a browser warning.
- **Domain** — checks the domain registration itself, from the registry. Catches the renewal everyone assumed somebody else had handled.
- **Heartbeat** — your job checks in with us and silence is the alarm. Catches a backup or cron job that quietly stopped, which nothing checking from outside can see.

Heartbeats are one line added to a job you already run:

    0 4 * * * /usr/local/bin/backup.sh && curl -fsS https://lastseenup.com/hb/<token>

That URL always answers OK — for an unknown token, an unreadable payload, or a
bad day on our end. A monitoring service that returns an error into your cron job
is a monitoring service that breaks your cron job.

## What it will not do

- **It will not wake you because something got slow.** Degraded shows on the dashboard and the chart, but never sends an alert. Only going down does.
- **It will not send twelve alerts for one bad night.** A monitor bouncing more than three times in half an hour is marked as flapping and collapsed into a single notification until things have been quiet for fifteen minutes.
- **It will not alert during your own deploy.** A maintenance window pauses alerts, not the record.
- **It will not invent an outage.** If the data cannot be trusted — a restart, a late write — nothing opens at all.

A failure is confirmed over consecutive checks before an incident opens; the
default is two. Recovery needs agreement from every region that is checking.

## How alerts reach you

Push notification to everyone on the team, plus an iOS Live Activity counting
the outage up on the lock screen. One tap acknowledges without unlocking. If
nobody acknowledges, it escalates: another push, then a phone call if one is
configured.

## What the dashboard holds

Every individual check for the last thirty days, hourly summaries kept
indefinitely, how fast each region usually answers and how slow it gets on a bad
day (p50 and p95), uptime across the last ninety days, and each incident with
its full timeline and who acknowledged it. It updates live as monitor state
changes.

## Pages

- [Home](https://lastseenup.com/)
- [iOS app on the App Store](https://apps.apple.com/us/app/last-seen-up/id6797648903)
- [Privacy policy](https://lastseenup.com/privacy)
- [Terms of service](https://lastseenup.com/terms)
- [Cookie policy](https://lastseenup.com/cookies)

## Contact

- Support: support@lastseenup.com
- Security: security@lastseenup.com
- Privacy: privacy@lastseenup.com
- Legal: legal@lastseenup.com