Downtime

A key metric that all monitors calculate is Downtime.

This is the amount of time that the monitored service was offline. Specifically, downtime is accrued while the monitor has an internal status of Down or Partial.

Dashboard

The easiest way to view downtime is on a monitor's dashboard. Here, the Downtime panel provides various metrics, calculated over a period of last hour, last 24 hours, last 7 days, last 30 days, and last 365 days.

If the monitor has existed for less than the length of a given period, then the calculation is performed starting when the monitor was created.

For example, if a monitor was created 2 hours ago, was Online for the first hour, but then Down for the second hour, the last hour availability would show as 0% and the last 24 hours availability (and that of all other periods) would show as 50%.

Total

This metric is the combined length of all downtime in the specified period.

That is, the total length of time in which the monitor had a state of Down or Partial.

Count

This is the number of discrete instances of downtime within the specified period.

The count does not increase when changing between Down and Partial.

For example:

• If a monitor was Down for the entire duration, the count would show 1, as only a single period of downtime occurred.
• If the monitor started Online, then was Down for a period, then recovered back to Online, the count would also show 1, as only a single period of downtime occurred.
• If it started Down, then went to Partial, then back to Down, then back to Partial, the count would still read 1, as (despite the status change) only a single period of downtime occurred.
• If it started Down, recovered to Online, but regressed back to Down, the count would read 2, as there were multiple discrete instances of downtime.

Longest

This value represents the longest discrete instance of downtime within the specified period.

As with the count, status changes between Down and Partial are ignored. An instance of downtime ends only when the internal status changes to another value.

Average

This is the simple arithmetic mean length of downtime within the specified period.

It is calculated as the total duration of all downtime, divided by the count of the number of discrete instances.

Availability

This metric shows the proportion of time during which the monitor was online.

Specifically, it is calculated as the total time in the specified period in which the internal status was Online or Degraded, divided by the total length of that period.

As with other metrics, if the monitor has existed for less than the length of a given period, then the period is truncated to start when the monitor was created.

Status Pages

By default, status pages show the availability metric for each monitor they list.

Other downtime metrics are not provided directly, but may be able to be determined depending on what other data is also shown (namely, incidents).

It is also possible to hide the availability metric, to only show the current status.