Alerts & Notifications

EmberNet provides a flexible alerting system that monitors metrics, evaluates conditions, and dispatches notifications to the appropriate channels when thresholds are breached.

Alert Concepts

Alert Rules

An alert rule defines:

• Metric — The data source to evaluate
• Condition — The threshold or expression that triggers the alert
• Duration — How long the condition must persist before the alert fires
• Severity — The alert severity level
• Channels — Where notifications are sent

Alert States

State	Description
OK	Metric is within normal bounds
Pending	Condition is met but duration threshold not yet reached
Firing	Alert is active and notifications have been dispatched
Acknowledged	An operator has acknowledged the alert
Resolved	Condition has returned to normal

Severity Levels

Severity	Description	Typical Response
Info	Informational, no immediate action required	Log and review
Warning	Potential issue, investigation recommended	Review within shift
Critical	Immediate action required	Immediate response

Creating Alert Rules

Basic Threshold Alert

1. Navigate to Alerts → New Rule
2. Select the metric to monitor
3. Define the threshold condition (e.g., Temperature > 185°F)
4. Set the evaluation duration (e.g., for 5 minutes)
5. Choose a severity level
6. Select notification channels
7. Save the rule

Example Alert Rules

Rule	Condition	Duration	Severity
High Temperature	temperature > 185	5 min	Critical
Low Disk Space	disk_usage_pct > 90	10 min	Warning
Device Offline	heartbeat = 0	2 min	Critical
High CPU Usage	cpu_usage > 95	15 min	Warning
Vibration Anomaly	vibration_rms > threshold	1 min	Critical

Notification Channels

Email

Send alert notifications to one or more email addresses:

• Configure SMTP server settings
• Define recipient lists per alert rule
• Customize email subject and body templates

Webhook

Dispatch HTTP requests to external systems:

• Configurable URL, method, headers, and payload
• Support for JSON and form-encoded payloads
• Retry logic for failed deliveries

SMS

Send text message notifications:

• Provider-agnostic SMS integration
• Phone number lists per alert rule
• Character-limited message templates

PagerDuty

Integrate with PagerDuty for on-call alerting:

• Automatic incident creation and resolution
• Severity mapping to PagerDuty urgency levels
• De-duplication of repeated alerts

Alert Management

Acknowledgment

When an alert fires, operators can acknowledge it to indicate that it is being investigated. Acknowledging an alert:

• Records the operator's name and timestamp
• Optionally suppresses further notifications for the duration
• Does not resolve the alert — it remains in the acknowledged state until the condition clears

Silencing

Temporarily suppress alerts during maintenance windows:

• Define silence rules by metric, device, or alert rule
• Set start and end times for the silence
• Silenced alerts are still evaluated but notifications are suppressed

Escalation

Configure escalation policies for unacknowledged alerts:

• Define escalation tiers with increasing severity
• Automatically escalate after configurable time periods
• Route escalated alerts to management or on-call rotations

Alert History

All alerts are logged with full history:

• Timestamp of state transitions (firing, acknowledged, resolved)
• Operator actions and notes
• Metric values at the time of the alert
• Notification delivery status

Access alert history from Alerts → History in the dashboard.

Next Steps

• Configuration Reference — Configure alert settings
• API Reference — Manage alerts via API
• FAQ — Common alerting questions