Maintenance and curation
Maintenance windows
Section titled “Maintenance windows”Announce a window, and for its duration no new incidents open and the time is SLA-neutral. The window still lands on the record and can show as a banner on your status page, with a visibility toggle per window. Planned work is visible, not counted against you, and never silently excluded.
With maintenance:write, the REST API creates, updates, cancels, and deletes
maintenance windows, so your deploy pipeline can announce its own work.
Outage curation
Section titled “Outage curation”Measured downtime is kept separate from the incident story. Discard a false positive with a reason, and Perstat previews its effect on uptime, before and after, ahead of applying it. Every discard is soft, carries the actor, and is reversible. A restored window is exactly the window that was discarded, on its real boundaries.
History
Section titled “History”Three tabs carry the full story, org-wide and per monitor, each one deep-linkable:
- Timeline: raw status changes per region, unedited.
- Outages: the curated downtime record, discards included and labeled.
- Activity: the configuration history, including “report discarded” and “report restored” with who and when.
Why corrections are visible
Section titled “Why corrections are visible”Every correction keeps its actor, reason, effect, and a way back, so reviewers see exactly what changed and why. How much weight a correction carries in an audit is for your auditor to decide. Evidence and retention covers the report rules and the history horizons.
- Evidence and retention: how curated records become SLA figures.
- Status pages: where maintenance banners appear.