Statuses

A site has one status at a time. The app, the status page and the uptime badge all read it from the same rule, so they can't disagree about a site. This page lists the statuses, the order in which the rules apply and what the public sees.

Statuses in the app

The rules apply from the top of the table down, and the first one that matches decides. A workspace whose monitoring is paused shows its sites as Paused. Only a few values are stored: a site is remembered as up, down or unknown, and the other statuses are worked out when you read the site.

Content changes are an event and not a status. A site that changed its page text in the last 24 hours shows a Changed chip next to its status, so a site can be Online and Changed at the same time.

When a site turns Offline

A failed check doesn't make the site Offline straight away. OctoPing first weighs it against the probe regions, as described on the vantage points page. A failure the probes back up, or that they can't speak to within the confirmation wait, turns the site Offline. A failure every probe contradicts leaves it up and shows Partial.

The status and the alert are two decisions. Offline starts at the first confirmed failure. The down alert waits until the notification's number of consecutive failures is reached (3 on the default email notification), and repeats no more often than the notification's repeat interval.

Heartbeat statuses

Heartbeats use their own four statuses. The same order applies: the first match wins.

Statuses on a public status page

A public page uses a shorter vocabulary. Visitors see whether something is available, so internal states such as Disabled don't appear by name.

Partial reads Operational on a public page because visitors can reach the site, and only OctoPing's checker was refused. On the page's 90-day bars the same days draw as successful days. Heartbeats map Healthy to Operational, Unhealthy to Down and Maintenance to Under maintenance.

The banner at the top of the page folds the item statuses into one, using the first rule that matches.

An item with the status Unknown doesn't change the banner, which treats it like Operational.

Incidents override monitoring

While an incident is open on a status page, the impact set for a component replaces the status monitoring would show for it. Degraded performance and partial outage show as Degraded, and major outage shows as Down. The banner is worked out after the override, so an incident that marks a component down outranks that component's maintenance.