Downtime operations

This section describes how to add/edit/delete downtime.

A downtime contains the following information: a schedule start time and a schedule end time.

Downtimes can be triggered to start or stop immediately.

Downtimes are organized using tags from the monitors.

Downtime types

Two schedule types are available (the Type selector in the downtime form):

Custom — a one-off maintenance window defined by a start and an end date/time (a date-time range picker). Once the end time is reached the downtime is done.
Regular — a recurring weekly window. Pick one or more start days of the week plus a start time, and one or more end days of the week plus an end time. The webserver turns this into internal cron schedules and re-triggers the downtime every week until you disable or delete it. Typical use: “every Saturday 22:00 to Sunday 06:00” for a weekly batch or backup window.

The type of an existing downtime cannot be changed while it is running.

Downtime scope: monitors, tags and applications

A downtime can target, in any combination:

explicit monitors of every type (browser scripts, URL, API, TCP, UDP, ping, nslookup, DB, SNMP devices, WMI devices, system/hosts),
tags per monitor type — every monitor of that type carrying the tag is included,
applications — all monitors that belong to the selected app(s) are included, child (nested) apps are expanded recursively.

While the downtime is active, the affected monitors are not scheduled to run and their status is frozen; when it ends, their pre-downtime status is restored.

Impact preview

Before saving, the downtime form shows an impact preview: the current scope (monitors + tags + apps) is resolved server-side — without saving anything — and the resulting list of affected monitors and applications is displayed, grouped by monitor type, with the probe each monitor runs on. Use it to verify a tag does not pull in more monitors than intended.

Downtimes are organized by tags

The computed list of monitors

Downtimes can be associated with tag(s) and/or monitor(s). The list of monitors they reference is generated from the tag(s) and/or monitor(s).

For instance: mon1 has tag1, mon2 has tag1 , mon3 has tag2, mon4 has no tag. A given report has monitor mon4 and has tag tag1, then the computed list of monitors for this report will be mon1, mon2, mon4. The downtime will be triggered for each monitor in this list.

Accessing downtimes and user privileges

Users in the ADMIN group can add, edit, or delete a downtime, as well as add or remove monitor(s) and/or tag(s) from a downtime.

UUsers with tags matching those on the downtime and belonging to the USER group can also add, edit, or delete a downtime, as well as add or remove monitor(s) and/or tag(s) from a downtime

A downtime can have one or more tags.

The downtimes page

Open downtime from the top menu (or #/downtime). Each row shows the downtime’s display name, description, date / time range, the monitors it covers (counted per type), its tags, the user who owns it and its state: on (scheduled or running), off (disabled) or done (a custom window that has ended).

The row actions are Edit (pencil), Force start (play), Force stop (stop), Disable / Enable (toggle) and Delete (trash).

Add a Downtime

Click + next to the search box.
In Settings, enter the Display Name, an optional description, and whether it is Enabled. Under Schedule, choose the Downtime Type (custom or regular, see Downtime types) and set its dates or weekly days and times. The small calendar under the picker shows the other downtimes of the month, so an overlap is visible before you save.
In Monitors, add the monitors, tags and applications it covers. Check the impact preview.
Click the disk icon to save. The downtime appears in the list and a notification confirms it.

Edit a Downtime

Click the pencil of the row. The same form opens with the downtime’s values; the header shows who changed it last. Change what you need and click the disk icon. The duplicate icon next to it saves the form as a new downtime instead.

Edit downtime form with its schedule and month calendar

Start a Downtime immediately

Click Force start (play). The downtime starts now, without waiting for its schedule: the monitors it covers stop being scheduled and their status is frozen. A notification confirms that the start was sent to the probes.

Stop a Downtime immediately

Click Force stop on a running downtime. It ends now, the monitors are scheduled again and their pre-downtime status is restored.

Delete a Downtime

Click the trash of the row and confirm with OK. A running downtime cannot be deleted: stop it first. To delete several downtimes, tick their rows and use the trash button of the toolbar.

Calendar view

Toggle between the default table view and a calendar view using the two buttons above the downtime list (table icon / calendar icon). The calendar view is useful for spotting overlapping maintenance windows at a glance.

Displays a month grid; navigate with the previous/next arrows.
Both regular (recurring weekly schedule) and custom (one-off date range) downtimes are expanded into concrete chips on the days they cover. A chip spanning multiple days is truncated with a continuation marker on the days it doesn’t start/end on.
Chip color distinguishes regular vs custom downtimes; a dimmed style marks a downtime that is off/done, and a highlighted style marks one currently running.
A day with more than one downtime active shows an overlap badge (e.g. 3×) in its corner.
Hovering a chip shows its full name, type, time range, state, and owning user.
Click a chip to open that downtime in the edit/view modal (same as the table’s pencil icon). Right-click a chip for a context menu with the same actions available in the table row (edit, enable/disable, force start/stop, delete).

Filter downtimes

Type in the Filter table box to match any column (name, description, monitors, tags, user).
Use the filters under the headers: text filters for display name, description, monitors and user, drop-downs for tags and state.
The page size selector and the arrows under the table paginate it; the refresh button reloads it.