Skip to main content

​
Overview

Dashboard alerts proactively notify you when metrics cross defined thresholds, ensuring you respond quickly to important changes without constantly monitoring screens. This guide covers creating and managing effective alert rules.

​
Alert types

The Dashboard supports multiple notification methods:

Email notifications

Detailed alert messages sent to your inbox with metric context and links.

In-app alerts

Real-time notifications displayed within the Dashboard interface.

Webhook alerts

Programmatic notifications sent to external systems like Slack or PagerDuty.

Mobile push

Notifications sent to your mobile device for critical issues.

​
What you will learn

This exploration task teaches you to:
  • Create alert rules based on metric thresholds
  • Configure notification channels
  • Set up alert severity levels
  • Manage alert fatigue with grouping and scheduling
  • Acknowledge and resolve alerts

​
Prerequisites

  • Completed Creating custom dashboard views
  • At least one custom dashboard with widgets configured
  • Valid email address or notification channel configured
  • Understanding of metrics you want to monitor
1

Access alert management

Navigate to “Alerts” from the main menu or click the alert bell icon in the Dashboard header.
The alerts page shows existing alert rules and recent alert history.
2

Create a new alert rule

Click “New alert” or “Create rule” to start configuring an alert.
Start with one critical metric to learn the process before creating multiple alerts.
3

Select the metric to monitor

Choose the metric and data source for your alert:
  • Browse available metrics from your dashboards
  • Select the specific widget or data stream
  • Define the time aggregation (average, sum, count, etc.)
4

Define threshold conditions

Set when the alert should trigger:
  • Above: Alert when metric exceeds value
  • Below: Alert when metric drops below value
  • Equals: Alert when metric matches specific value
  • Anomaly: Alert on unusual patterns (requires historical data)
Choose thresholds based on actual business impact, not just statistical limits.
5

Set evaluation window

Configure how often the metric is checked:
  • Real-time: Check every minute (for critical metrics)
  • Standard: Check every 5 minutes (for most use cases)
  • Hourly: Check once per hour (for slowly changing metrics)
Too frequent evaluation can cause alert fatigue. Too infrequent may miss important changes.
6

Configure notification channels

Select how you want to be notified:
  • Email addresses
  • Webhook endpoints
  • In-app notification settings
  • Mobile push (if enabled)
7

Set severity level

Assign a severity to help prioritize responses:
  • Critical: Immediate action required (system down, data loss)
  • Warning: Attention needed soon (approaching limits, performance degradation)
  • Info: Awareness only (updates, routine changes)
8

Add alert details

Provide context to help responders understand the alert:
  • Descriptive name
  • Explanation of what the metric means
  • Suggested actions or runbook links
  • Escalation contacts
9

Test the alert

Use the “Test” button to verify the alert configuration works correctly.
You should receive a test notification through your configured channels.
10

Enable the alert

Toggle the alert to “Active” to start monitoring.

​
Alert best practices

Well-designed alerts improve response times without overwhelming your team.

​
Threshold selection

  • Set thresholds that represent actual problems, not just statistical anomalies
  • Use historical data to establish baseline ranges
  • Build in buffer zones to avoid flapping (rapid on/off alerts)
  • Review and adjust thresholds quarterly

​
Reducing alert fatigue

  • Group related alerts: Combine multiple conditions into single notifications
  • Set quiet hours: Pause non-critical alerts during maintenance windows
  • Use auto-resolution: Clear alerts automatically when metrics return to normal
  • Implement escalation: Route unacknowledged alerts to managers after time periods

​
Alert maintenance

Regularly review your alerts:
  • Disable alerts that fire frequently but require no action
  • Update thresholds as business needs change
  • Remove alerts for deprecated metrics
  • Document alert purposes for team knowledge sharing

​
Managing active alerts

When an alert fires:
  1. Acknowledge: Click “Acknowledge” to indicate you are handling the issue
  2. Investigate: Use linked dashboards to understand the root cause
  3. Resolve: Mark the alert as resolved once the issue is fixed
  4. Document: Add notes about what happened and how it was resolved

​
Troubleshooting

  • Verify the alert is enabled
  • Check if the metric is reporting data
  • Confirm threshold conditions match your expectations
  • Review evaluation window settings
  • Adjust thresholds to be less sensitive
  • Increase evaluation windows
  • Group multiple conditions
  • Implement quiet hours for non-critical times
  • Check notification channel configuration (email, webhook URLs)
  • Verify notification preferences in your profile
  • Check spam/junk folders for email alerts
  • Test notification channels individually

​
Next steps

With alerts configured, learn to Analyzing dashboard data to better understand trends and patterns in your metrics.