One recent addition to the Vigil service is uptime calculation. We've had a few questions about it, so thought we'd provide a little more explanation of how we arrive at this number for your site. The uptime calculation is based on the last 30-days of data we have on a site.
The calculation is based on the fraction of the time a site check came back okay versus the total number of checks we ran on a site. The amount of time between checks is also taken into account. For example, if a site check fails, it will be considered down until the next check that returns successfully. The granularity of the uptime report will vary depending on the plan you are on, and how often we check your sites.
For the weekly report, the uptime calculation is performed when the report itself is generated. Otherwise, we calculate the uptime every hour, and this is the value presented in the client.
You may see that your uptime is not 100%, but you never received a down notification. Depending on how you have your notifications configured, we may perform more than one check before we send a notification to let you know your site is down. This is to avoid false positives, and too many alerts when there may only be an intermittent network issue. These individual failures, while not triggering an alert, will still be used in a site's total uptime calculation. For example, the default alert sensitivity requires 3 failed checks before we send an alert. If a site only failed two checks but then came back, it would never send out an alert. Those two failed checks would still drop the uptime from 100%.
If you have any questions about how uptime is working with your sites, please let us know.