For the complete documentation index, see llms.txt. This page is also available as Markdown.

Configure Alert Rules

Learn how to configure alert rules in Harness AI SRE to route, filter, and enrich incoming alerts.

Alert rules define how incoming alerts are processed and when incidents should be created.

Overview

Alert rules help you:

  • Create incidents from incoming alerts

  • Map alert data to incident fields

  • Set up automated responses with conditional logic

  • Configure Harness AI SRE on-call notifications and paging

  • Associate relevant runbooks for automated response


Alert rule capabilities

Field-based conditions

The default mode for alert rules uses a visual builder where you can:

  • Select alert fields from your monitoring integrations

  • Choose comparison operators (equals, contains, greater than, and others)

  • Combine multiple conditions with AND/OR logic

  • No coding required

CEL expression conditions

Write advanced boolean expressions for complex filtering:

  • Regex pattern matching for service names

  • Complex multi-field logic

  • String operations and numeric comparisons

  • More concise than multiple field-based conditions

Go to Use CEL to Route Alerts to understand CEL expression mode.

Incident creation

Automatically create incidents from matching alerts:

  • Map alert fields to incident properties

  • Set incident severity and priority

  • Associate services and environments

  • Populate custom fields

On-call notifications

Page the on-call team when critical alerts arrive:

  • Select impacted services

  • Configure escalation policies

  • Choose notification channels

  • Set response time expectations

Runbook association

Attach automated response procedures:

  • Automatically trigger runbooks on incident creation

  • Suggest runbooks to responders

  • Provide reference documentation


Getting started

Go to Create an Alert Rule for step-by-step configuration instructions.


Best practices

Keep these guidelines in mind when you create and maintain alert rules:

  • Create specific rules for different types of alerts

  • Use clear, descriptive names for your rules

  • Test rules with sample alerts before activating

  • Review and update rules as your services evolve

  • Document your rule configurations for team reference


Next steps

Last updated

Was this helpful?