Status page
A public, read-only page your users can check before they email you: each check's uptime, the incidents you post, and your announcements, such as planned maintenance. It's built from your uptime checks.
Find it in the Monitoring tab, under Uptime, with Customize Status Page.
Page settings
Public status pageon or offPublishes the page at /status/YOUR_PROJECT_ID on the dashboard's address, or at your own domain. While it's off, visitors get a 404.
Publishing makes check names and uptime history public, along with your incidents and announcements, so name your checks with that in mind. Unpublishing keeps your checks running, and your incidents and announcements for when you publish again.
Outage notice on your docson or offShows a short notice on your published documentation while checks are failing, with no check names and no error details. It clears on its own once the checks recover, and links to the status page when that's published.
Custom domainhostnameServes the page from your own address, such as status.acme.com. Add it, create the two DNS records shown, a CNAME that routes traffic and a TXT that proves you own the domain, then press Verify. If you remove it, the page stays at its default address.
Each day's bar shows up at 99.9% or better, degraded at 90% or better, and down below that.
Incidents
An incident is your public account of what went wrong: a headline, then a timeline of updates until it's resolved.
TitletextrequiredThe headline, such as "Elevated error rates on /v1/orders".
First updatetextrequiredWhat you know right now. It starts the timeline.
StatuschoiceWhere it stands. Each later update can move it on.
| Value | Means |
|---|---|
investigating | You're looking into it |
identified | You know the cause |
monitoring | A fix is out and you're watching it |
resolved | It's over. The incident moves into the history |
ImpactchoiceHow bad it is, which sets its colour on the page: minor, major, critical, or maintenance for planned work.
Affected checkslist of checksThe checks it affects, shown as chips on the incident.
Post updates as you learn more, each with its own status, then resolve it. Every update is signed with the poster's display name, never their email.
An incident is your public record of what happened. Deleting one deletes every update posted to it.
Announcements
Notices on the page for what your uptime graph won't say: maintenance, a migration, a deprecation. They show whether or not anything is down.
TitletextrequiredThe headline, such as "Scheduled maintenance on Sunday".
DetailtextMore text under the title.
LevelchoiceIts colour: info, success, warning or critical.
Startsdate and timeWhen it starts showing. Empty: straight away.
Endsdate and timeWhen it stops. Empty: until you remove it.
From the CLI
Post and update incidents with octri monitoring incidents, and manage banners with octri monitoring announcements.
Related
- Uptime checks: what the page is built from.
- Alerts: be told when a check goes down.
- MCP outage warnings: agents are told about an outage from the same checks.