Setting Up Alarms

This guide walks you through creating alarm rules that automatically trigger when device telemetry crosses your defined thresholds.

Prerequisites

At least one device sending telemetry data
ADMIN or ENGINEER role
Know which telemetry fields and thresholds matter for your use case

Steps

Identify Critical Fields

Review your device telemetry to decide what to monitor. Common examples:

Field	Condition	Severity
`temperature`	Above 80°C	CRITICAL
`battery`	Below 20%	WARNING
`humidity`	Above 90%	INFO
`pressure`	Below 950 hPa	WARNING

Create an Alarm Rule

Navigate to Alarms and click Create Alarm Rule.

Fill in the rule configuration:

Name: Descriptive name (e.g., "High Temperature Alert")
Application: Select the target application
Device (optional): Target a specific device, or leave blank for all devices in the application
Telemetry Field: The field to monitor (e.g., temperature)
Condition: The comparison operator (>, >=, <, <=, ==, !=)
Threshold: The trigger value (e.g., 80)
Severity: INFO, WARNING, or CRITICAL

Set a Cooldown Period

The cooldown prevents alarm flooding by limiting how often the same rule can trigger.

Example: A 5-minute cooldown means that after triggering, the rule won't fire again for 5 minutes even if the condition remains true
Set shorter cooldowns for critical alarms (1–2 minutes)
Set longer cooldowns for informational alerts (15–30 minutes)

Enable the Rule

Toggle the rule status to Enabled. The platform immediately starts evaluating incoming telemetry against your rule.

Rules are evaluated in real time as telemetry arrives. There is no polling interval — alarms trigger as soon as the condition is met.

Test with Sample Data

Send telemetry that exceeds your threshold to verify the alarm works:

mosquitto_pub \
  -h broker.sensorclouds.io \
  -u "device-username" \
  -P "device-password" \
  -i "device-client-id" \
  -t "v1/devices/telemetry" \
  -m '{"temperature": 85}'

You should see:

A new alarm instance appear in the Alarm Instances list
A notification in the Notification Center
A real-time WebSocket event (if viewing the dashboard)

Monitor Alarm Instances

Go to Alarms → Instances to see all triggered alarms. Each instance shows:

The rule that triggered
Which device triggered it
The telemetry value that crossed the threshold
When it triggered
Current status (ACTIVE, ACKNOWLEDGED, RESOLVED)

You can acknowledge alarms to indicate someone is investigating, or resolve them once the issue is fixed.

Best Practices

Start with fewer rules: Add alarms for the most critical conditions first, then expand
Use severity levels meaningfully: Reserve CRITICAL for conditions requiring immediate action
Set appropriate cooldowns: Too short causes alert fatigue; too long risks missing repeated issues
Name rules clearly: Include the field and condition in the name (e.g., "Battery Below 20%")
Review alarm history: Periodically check which alarms fire most often to tune thresholds

On this page