This page is for operators: how to find alarms, take them, close them and leave a trail. For what an alarm is, see Alarms.
The Alarms page
Open Alarms in the IoT menu. The subtitle reads “Alarms of your devices: acknowledge, clear and follow the runbook.”
| Element | What it does |
|---|---|
| Open (n), Acknowledged (n) | Counts of the active alarms. |
| Status | Not resolved (default: open and acknowledged), Open, Acknowledged, Resolved, All statuses (last 7 days). |
| Severity | One or more of the five severities. |
| Search alarms | Searches titles, types and device names. |
| Columns | Severity, Alarm, Device, Status, Opened. Sort by one column at a time; the column chooser is remembered per browser. |
| Rows per page | 20 by default. |
| Gear icon | Alarm settings (administrators). See Alarm notifications in depth. |
Filters, sort and page live in the page address, so you can bookmark or share a filtered list, for example …/admin/iot/alarms?status=OPEN&severity=CRITICAL,MAJOR.
Resolved and All statuses read the last 7 days of opened alarms. The API reads up to 92 days per request (see Alarms API).
New alarms arrive live; a pill “n new alarms: show” brings them into the list. Without a live connection the list refreshes every 30 seconds.
The device page’s Alarms tab shows the same list for one device.
Acknowledge or clear several alarms
- Tick the rows.
- Click Acknowledge (n) or Clear (n).
- Optionally write one note for all of them (up to 2,000 characters) and confirm.
The result reads “done done, skipped skipped”. An alarm that was acknowledged or resolved meanwhile is skipped, not failed. At most 100 alarms are changed per request.
Acknowledge all (n) acknowledges every open alarm, after asking, with one optional note.
Clear needs Resolve (clear) alarms by hand (Admin by default). Acknowledging needs Acknowledge alarms and add notes (User by default).
The alarm page
Click an alarm to open it. The header shows the severity, the status, the title, Device (a link to the device), Opened time, the Type, the Topic when the rule filters by one, and these marks when they apply:
| Mark | Meaning |
|---|---|
| Fired n times | The condition fired again while the alarm was active. |
| Flapping | The alarm opened and cleared too often; its clear is held. |
| Backfilled | Opened from readings that arrived late. |
Actions
| Button | Does |
|---|---|
| Acknowledge | Takes the alarm. The status becomes Acknowledged. |
| Clear | Resolves it by hand. If the condition is still true, the rule does not reopen it until the condition has been false once. |
| More actions > Acknowledge with a note… / Clear with a note… | Asks “What did you do?” with Note (optional), up to 2,000 characters. |
If someone else changed the alarm first, you see “Nothing changed: the alarm was already status.”
Runbook
The Runbook box shows the rule’s runbook text. If the rule has none, a generic one appears: check the device is reachable, look at its latest values and recent connection events, fix the cause, then clear the alarm with a note.
Related
Raised by names the source: Profile rule name (default flow, alarm evaluation), Flow flow → block block, or An operator.
Timeline
The timeline lists every step, newest first, with who did it (you, an operator, a flow, the platform):
| Line | Meaning |
|---|---|
| Alarm opened | |
| Severity changed: from → to | |
| Acknowledged, Cleared | With the note, if any. |
| Notified | A notification went out (how many recipients, which channels). |
| Notification held back (throttle) | The same change was told less than the repeat window ago, or went into a storm digest. |
| Muted: nobody was notified | The device was muted. |
| Re-open held back after a manual clear | |
| Flapping: clears held back | |
| Automation | Changed by a flow or rule. |
| Note | A note someone added. |
| “from late readings” | Added to steps caused by late readings. |
To add a note without changing the alarm, write it in Write a note for the timeline and click Add note. Notes are limited to 2,000 characters and cannot be empty (“A note needs some text.”).
Good practice
- Acknowledge when you start working on an alarm, so reminders stop and colleagues see it is taken.
- Clear with a note that says what you did. The note stays with the alarm for its whole history.
- Because alarms clear themselves when the condition is over, review Resolved alarms that were never acknowledged, for example once per shift.
- Use the bell for a quick look and to acknowledge open alarms inline. See The bell.