Components
A component is a part of your service that customers see on a status page, for example "API", "Website" or "Checkout". Spectate sets the status of each component from the incidents and maintenance that you publish. You do not set a component's status by hand.
Add a component
- In the dashboard, open the status page and go to the Components tab.
- Select Add component.
- Type a Name, for example
API. - Optional: type a Group, for example
Core services. - Optional: type a Description. Visitors see it on the public page.
- Optional: select a Linked monitor.
- Select Create component.
New components go to the end of the list. Use Move up and Move down on a component to change the order. Use Edit to change a component and Delete to remove it.
Groups
A group is a heading with components under it. To put a component in a group, type the group name in the Group field. Components without a group show under Ungrouped in the dashboard.
Each group has a default state on the public page: Expanded or Collapsed. Visitors can open and close a group. The default state only sets how the group shows when the page loads.
Component statuses
| Status | Value | Public label | Meaning |
|---|---|---|---|
| Operational | operational | Operational | Works normally. |
| Degraded | degraded_performance | Degraded performance | Slower or less reliable than usual. |
| Partial outage | partial_outage | Partial outage | Down for some visitors or some functions. |
| Full outage | major_outage | Full outage | Down. |
| Maintenance | under_maintenance | Maintenance | In planned maintenance. |
How Spectate sets the status
A component's status is the worst of these, in this order:
- The worst impact on the component from open incidents. An incident impact is
degraded_performance,partial_outageormajor_outage. under_maintenance, if a maintenance that affects the component is in progress.operational, if nothing else applies.
Any incident impact is worse than maintenance. Spectate calculates the status again each time an incident, an update, an impact period or a maintenance changes. When you resolve an incident, its components go back to operational, unless another incident or maintenance still affects them.
The headline of the page is the worst status of all components. On the Components tab, each component shows its status and the incident or maintenance that causes it, for example "Partial outage · from API errors".
Change a component's status
You cannot set a status on a component directly. To change it, do one of these:
- Publish an incident, or post an update, with an impact on the component.
- Edit the impact timeline of an incident. See incidents.
- Schedule a maintenance that affects the component.
Link a monitor
Link a component to an uptime monitor to get measured uptime and latency, and to let the monitor open incidents. Select the monitor in the Linked monitor field. You can link one monitor per component, from the same team. You cannot link a host.
A linked monitor gives the component these things:
- Measured uptime. The uptime bars use the checks of the monitor. A day's colour is the worse of the measured downtime and the incident impacts of that day.
- A latency chart. Turn on Show primary-region latency chart on the component. The chart shows the response time of the last 24 hours. It only shows for components with a linked monitor.
- Automatic incidents. These need the automation setting of the page. See below.
A linked monitor does not change the component's status by itself. Only incidents and maintenance change the status.
Components without a monitor also have an uptime history. Spectate calculates it from the incident impacts. partial_outage and major_outage count as down. degraded_performance and maintenance count as up.
Automatic incidents from monitors
With automation, a monitor opens and resolves a public incident for each component that it is linked to.
- Go to Settings → Automation on the status page.
- Select Open and resolve incidents from linked monitors.
- Optional: type a Down message and a Recovered message.
- Select Save automation.
Automation is off by default. When it is on, Spectate does these things:
- The monitor goes down. When the downtime alert of the monitor fires, Spectate opens an incident "
<component>is down" with the statusinvestigating. The component getsmajor_outagefrom the time the monitor went down. - The monitor recovers. When the alert resolves, Spectate resolves the incident at the recovery time, with the recovered message.
- The monitor fails again within 30 minutes. Spectate opens the same incident again, not a new one.
- Another open incident already affects the component. Spectate does not open a second incident.
Each component gets its own incident, also when several monitors fail at the same time. If the alert opened a team incident, Spectate links the public incident to it. Spectate emails subscribers about automatic incidents as for incidents that you publish.
Empty messages use the defaults:
| Message | Default |
|---|---|
| Down message | We are investigating an outage affecting {component}. |
| Recovered message | {component} has recovered and is operating normally. |
In all messages, Spectate replaces {component} with the component name and {status_page} with the page name.
Take over an automatic incident
The dashboard marks an automatic incident with Automatic. If a person posts an update or edits the impact timeline, the incident is no longer automatic. After that, automation does not change or resolve it.