Monitors overview

A monitor is a single thing you want Aloft to watch — a website, an API endpoint, a TCP service, a host, a TLS certificate, a domain registration, or a background job. Aloft checks each monitor on its own schedule, records every result, and opens an incident (and notifies your alert channels) when something goes wrong.

You’ll find all of your monitors under Monitors in the top navigation. This page is the home base for everything related to watching your endpoints.

Monitor types

Aloft supports seven monitor types. Pick the one that matches what you’re watching:

• HTTP / HTTPS — request a URL and check the status code. See HTTP and keyword monitors.
• Keyword — request a URL and check that a string is present (or absent) in the response body. See HTTP and keyword monitors.
• TCP port — open a socket to a host and port to confirm a service is accepting connections. See Port and ping monitors.
• Ping — send an ICMP echo to a host or IP to confirm it’s reachable. See Port and ping monitors.
• SSL certificate — read a host’s TLS certificate and alert before it expires. See SSL and domain expiry monitors.
• Domain expiry — look up a domain’s registration and alert before it expires. See SSL and domain expiry monitors.
• Heartbeat — wait for your job to call in; alert when the expected check-in doesn’t arrive. See Heartbeat monitors.

Monitor statuses

Every monitor carries a status that the worker keeps up to date:

• Up — the most recent check succeeded.
• Down — checks are failing. A monitor only flips to Down after enough consecutive failed checks (see Scheduling and confirmations), so a single blip doesn’t cry wolf.
• Degraded — the check succeeded, but the response was slower than the response-time threshold you set. The endpoint is reachable, just sluggish. Degraded still counts as “up” for uptime math.
• Paused — you’ve stopped checking this monitor. No checks run and no alerts fire until you resume it.
• Unknown — Aloft hasn’t recorded a result yet. New monitors start here until their first check lands. (On the dashboard’s status filter this is labeled Pending.)

A tour of the dashboard

The Monitors page lists every monitor in your organization, with summary tiles across the top (Total, Up, Down, Paused). Above the list is a filter bar:

• Status chips — click a status (All, Up, Down, Degraded, Paused, Pending) to show only monitors in that state.
• Search — type into the search box to filter by monitor name or URL, then press Search.
• Tag chips — if any of your monitors have tags, a row of tag chips appears. Click one to show only monitors with that tag. See Tags and organizing.

Use the Clear button to drop all active filters at once.

Note

The dashboard watches the background worker too. If the worker that runs your checks stops reporting in, a “Worker is not reporting” card appears at the top of the page with how long it’s been since the worker was last seen. If you see it, your checks may be stalled — start or restart the worker process.

If there are any open incidents, an Ongoing incidents card lists the affected monitors and how long each has been down.

Pausing, resuming, and running a check now

Open any monitor from the list to reach its detail page. From the buttons at the top you can:

• Run check now — trigger an immediate check instead of waiting for the next scheduled one. Handy right after you create or fix something.
• Pause / Resume — stop or restart scheduled checks. A paused monitor never opens incidents or sends alerts.
• Edit — change any of the monitor’s settings.
• Delete — remove the monitor and its history.

Tip

Pause monitors during planned work to avoid noisy alerts. For recurring or scheduled downtime, set up a maintenance window instead (see Maintenance in the top navigation) so suppression happens automatically.

The detail page also shows uptime over the last 24 hours, 7 days, and 30 days, an average response time, a response-time chart, a 90-day uptime bar strip, and a log of past incidents.

Where to go next

• Scheduling and confirmations — set the check interval, timeout, and how many failures it takes to declare a monitor down.
• Tags and organizing — group and filter monitors with tags.