Monitor a Site

Monitoring starts with choosing the target. Websites and services are the obvious case, but campaigns and embeds count too: a new social campaign across Instagram and Facebook, or a site that plays hosted YouTube videos. latency covers both.

Every monitor can link an optional service whose issues also raise alerts. Those alerts come from official provider status pages and our own probes.

Monitors consist of several parts (tasks, schedules, alerts), which together form each monitor's configuration.

A basic website monitor from the CLI:

$ latency monitor https://google.com

This creates a TCP monitor for the Google homepage on port 443.

Bare IP addresses work too:

$ latency monitor 1.1.1.1

The monitor command picks a check from the input: IP addresses default to ICMP, hostnames to TCP.

A flag overrides the default:

$ latency monitor https://google.com --with.icmp

Or a specific port on an IP with ICMP:

$ latency monitor 1.1.1.1:443 --with.icmp

The --with.* flags add sub-tasks that run alongside the monitor.

Tasks lists the available sub-tasks and how to build custom ones.

Monitors run once a minute by default; the interval is configurable:

$ latency monitor 1.1.1.1 --interval=120
# can use m, hr, d, wk, mo, y
$ latency monitor 1.1.1.1 --interval=30m

Signed-in sessions can also set an alert contact when adding a monitor:

$