Uptime Kuma
gunkanmaki.uptime-kuma
A Docker Compose file for Uptime Kuma alone, generated from the catalog and checked by the safety inspector. No account needed.
Self-hosted uptime monitor that checks HTTP endpoints, TCP ports, DNS records, certificates and container health on a schedule, with public status pages and a wide set of notification integrations.
Uptime Kuma runs periodic checks of several kinds — HTTP with keyword or status matching, TCP port, ping, DNS, database query, certificate expiry and passive push endpoints — recording each result into a history rendered as availability charts. Failures trigger notifications through providers configured per monitor, with retry counts and resend intervals to suppress noise from transient blips. Public status pages can be published from a chosen subset of monitors, grouped and branded, without exposing the administration interface. A first deployment claims the administrator account on first visit and adds one HTTP monitor.
You know it worked when
- The interface loads and the administrator account is created.
- An HTTP monitor reports its target as up.
- Stopping that target causes the monitor to report down within its configured interval.
- A configured notification provider delivers a test message.
- A published status page shows the selected monitors without requiring a login.
Known sharp edges
- Monitor definitions and the full check history live in an embedded database inside the data directory, which must be a persistent volume.
- The first person to reach the interface creates the administrator account, so an instance left reachable before setup can be claimed by someone else.
- Checks depend on the container's own DNS and egress, so a monitor can report a failure that is really a network problem inside the deployment.
- History grows continuously with check frequency and monitor count, and retention has to be set deliberately rather than assumed.