Documentation › Alert and act
Scheduled downtimes
During a downtime, HYPHA keeps evaluating and recording, but notifies no one.
What a downtime does
- Covered monitors are evaluated and their states recorded; no notification goes out.
- The groups concerned carry the "downtime" mark in the Problems view.
- At the end, a group still non-OK notifies normally, once.
- An event marks the start and the end.
- On a status page, a component in downtime shows it.
Scope
- Monitors: a list, or all of them.
- Tags: the group or the host must carry all the given tags (
env:preprod,role:db). - Groups: specific groups (
host:web-02).
Duration and recurrence
- Once: now for a duration (1 h, 4 h, 1 day, or custom), or between two dates.
- Every day, every week (chosen days) or every month (day of the month), at a given time, for a duration, in the chosen time zone, until an optional end date.
Where to set it
- Downtimes page: list (active ones first), creation with a preview of the monitors affected.
- A host page: Put in downtime (1 h, 4 h, 1 day, custom).
- A monitor page, and a line of the Problems view.
- Through the API:
POST /api/v1/downtimeswith{"duration_s": 3600, …}starts right away (see API).
Updated: 19 November 2026