Alert Escalation Policy Construct

Learn how to configure alert escalation policies with the Checkly CLI.

Learn more about alert escalation policies in the alerts configuration.

Use alert escalation policies to control when and how often you receive alerts when checks start failing, degrade, or recover. This helps you fine-tune your alerting to reduce noise and ensure timely alerts.

Run-Based Escalation

import { AlertEscalationBuilder, ApiCheck } from "checkly/constructs"

new ApiCheck("run-based-alert-check", {
  name: "Check With Run-Based Escalation",
  alertEscalationPolicy: AlertEscalationBuilder.runBasedEscalation(
    2, // Alert after 2 consecutive failures
    { interval: 5, amount: 2 }, // Send 2 reminders, 5 minutes apart
    { enabled: true, percentage: 50 } // Alert if 50% of parallel runs fail
  ),
  request: {
    method: "GET",
    url: "https://api.example.com/health",
  },
})

Configuration

Configure your alert escalation policies using the AlertEscalationBuilder, which provides helper methods for the different escalation strategies:

Method	Description	Parameters
runBasedEscalation()	Alert after N consecutive failed runs	(failedRuns, reminders, parallelRunFailureThreshold)
timeBasedEscalation()	Alert after N minutes of continuous failures	(minutesFailing, reminders, parallelRunFailureThreshold)

Alert Escalation Builder Methods

runBasedEscalation · function · required

Creates an alert escalation policy that triggers after a specified number of consecutive failed runs.

Usage:

AlertEscalationBuilder.runBasedEscalation(
  failedRuns,
  reminders,
  parallelRunFailureThreshold
)

Parameters:

Parameter	Type	Required	Default	Description
failedRuns	number	✅	1	Number of consecutive failed runs before alerting
reminders	object	❌	-	Reminder notification configuration
parallelRunFailureThreshold	object	❌	-	Parallel run failure percentage threshold

Examples:

Immediate Alert

// Alert immediately on first failure
const immediateAlert = AlertEscalationBuilder.runBasedEscalation(
  1, // Alert after 1 failed run
  { interval: 5, amount: 0 }, // No reminders
)

new ApiCheck("critical-service-check", {
  name: "Critical Service Check",
  alertEscalationPolicy: immediateAlert,
  request: {
    method: "GET",
    url: "https://api.example.com/critical",
  },
})

Use cases: Immediate alerting and noise reduction for flaky services.

timeBasedEscalation · function · required

Creates an alert escalation policy that triggers after a specified duration of continuous failures.

Usage:

AlertEscalationBuilder.timeBasedEscalation(
  minutesFailing,
  reminders,
  parallelRunFailureThreshold
)

Parameters:

Parameter	Type	Required	Description
minutesFailing	number	✅	Minutes of continuous failures before alerting
reminders	object	❌	Reminder notification configuration
parallelRunFailureThreshold	object	❌	Parallel run failure percentage threshold

Examples:

Alert after 15 minutes

// Wait for batch jobs or slow services
const timeBasedAlert = AlertEscalationBuilder.timeBasedEscalation(
  15, // Alert after 15 minutes of failures
  { interval: 10, amount: 3 } // 3 reminders, 10 minutes apart
)

new ApiCheck("batch-job-check", {
  name: "Batch Job Monitoring",
  frequency: Frequency.EVERY_5M,
  alertEscalationPolicy: timeBasedAlert,
  request: {
    method: "GET",
    url: "https://api.example.com/batch-status",
  },
})

Use cases: Batch job monitoring, slow services, time-sensitive operations.

reminders · object

Configure reminder notifications while an alert is active.

Usage:

reminders: {
  interval: 10, // Send reminder every 10 minutes
  amount: 5     // Send up to 5 reminders
}

Parameters:

Parameter	Type	Required	Default	Description
interval	number	❌	5	Minutes between reminder notifications: 5, 10, 15, 30
amount	number	❌	0	Number of reminder notifications to send: 0, 1, 2, 3, 4, 5, 100000

Examples:

No Reminders

// Alert once, no reminders
reminders: {
  interval: 5,
  amount: 0
}

Use cases: Alert fatigue prevention, escalation management, notification frequency control.

parallelRunFailureThreshold · object

Configure threshold for checks running in parallel across multiple locations.

Usage:

parallelRunFailureThreshold: {
  enabled: true,
  percentage: 60 // Alert if 60% of parallel runs fail
}

Parameters:

Parameter	Type	Required	Default	Description
enabled	boolean	✅	false	Whether to use parallel run failure threshold
percentage	number	❌	10	Percentage of parallel runs that must fail (10-100, in increments of 10)

Examples:

Regional Redundancy

// Only alert if majority of regions fail
const regionalAlert = AlertEscalationBuilder.runBasedEscalation(
  2, // After 2 consecutive failures
  { interval: 10, amount: 3 },
  { enabled: true, percentage: 70 } // 70% of regions must fail
)

new ApiCheck("global-service-check", {
  name: "Global Service Check",
  locations: ["us-east-1", "us-west-2", "eu-west-1", "ap-southeast-1"],
  runParallel: true,
  alertEscalationPolicy: regionalAlert,
  request: {
    method: "GET",
    url: "https://global-api.example.com/health",
  },
})

Use cases: Multi-region monitoring, CDN availability, redundant service checking.

Examples

Group-Level Policy

import {
  AlertEscalationBuilder,
  ApiCheck,
  CheckGroupV2,
} from "checkly/constructs"

const groupAlertPolicy = AlertEscalationBuilder.runBasedEscalation(
  3, // Alert after 3 consecutive failures
  { interval: 15, amount: 3 }, // 3 reminders, 15 minutes apart
  { enabled: true, percentage: 60 }
)

const monitoringGroup = new CheckGroupV2("monitoring-group", {
  name: "Production Monitoring",
  alertEscalationPolicy: groupAlertPolicy, // Applies to all checks in group
  locations: ["us-east-1", "eu-west-1"],
  runParallel: true,
})

new ApiCheck("group-api-check", {
  name: "API Check with Group Alert Policy",
  group: monitoringGroup, // Inherits alert policy from group
  request: {
    method: "GET",
    url: "https://api.example.com/endpoint",
  },
})

Reminder Limits: Be cautious with reminder settings. Too many reminders can lead to alert fatigue, while too few might cause important issues to be overlooked.

Parallel Run Thresholds: Only use parallel run failure thresholds for checks that run in multiple locations simultaneously (runParallel: true).