Synthetic monitoring for critical user paths

Setting Up Synthetic Monitoring (Regular Critical Path Checks)

Synthetic Monitoring—simulating real user actions on schedule. Script opens browser, goes through scenario (e.g., checkout), records: does each step work, how long does it take, where are errors? This detects problems before users notice.

Difference from Uptime Monitoring

Simple ping to GET / — not synthetic monitoring. It checks server responds. Synthetic monitoring checks business functions work:

• User can register
• Search returns results
• Add to cart works
• Payment form opens

Critical Paths for Monitoring

Priority—paths critical for revenue and retention:

E-commerce:

1. Search product → product page → add to cart → start checkout
2. Login → account → order history
3. User registration

SaaS application:

1. Login → dashboard → key feature
2. Create new object (project/document/task)
3. API endpoints (for B2B clients)

Content site:

1. Search → results → article
2. Newsletter signup form
3. Contact form

Playwright-based Synthetic Monitoring with Checkly

// checkly: checkout-flow.spec.js
const { chromium } = require('playwright')
const { expect } = require('@playwright/test')

async function checkoutFlow() {
  const browser = await chromium.launch()
  const page = await browser.newPage()

  try {
    // Step 1: Open catalog
    await page.goto('https://example.com/catalog')
    await expect(page.locator('.product-grid')).toBeVisible()

    // Step 2: Open first product
    await page.locator('.product-card').first().click()
    await expect(page.locator('[data-testid="product-title"]')).toBeVisible()

    // Step 3: Add to cart
    await page.locator('[data-testid="add-to-cart"]').click()
    await expect(page.locator('[data-testid="cart-count"]')).toContainText('1')

    // Step 4: Go to cart
    await page.goto('https://example.com/cart')
    await expect(page.locator('.cart-items')).toContainText('1 item')

    console.log('Checkout flow: PASS')
  } finally {
    await browser.close()
  }
}

Datadog Synthetic Tests

# Create Synthetic Browser Test via API
import requests

test_config = {
    "name": "Checkout Flow - Production",
    "type": "browser",
    "config": {
        "request": {
            "url": "https://example.com",
            "method": "GET"
        }
    },
    "options": {
        "tick_every": 300,  # Every 5 minutes
        "min_failure_duration": 120,
        "min_location_failed": 2,
        "retry": {"count": 2, "interval": 300}
    },
    "locations": [
        "aws:eu-west-1",
        "aws:us-east-1",
        "aws:ap-northeast-1"
    ],
    "status": "live",
    "tags": ["team:frontend", "env:production"]
}

requests.post(
    "https://api.datadoghq.com/api/v1/synthetics/tests/browser",
    headers={"DD-API-KEY": API_KEY, "DD-APPLICATION-KEY": APP_KEY},
    json=test_config
)

Playwright + GitHub Actions: Self-Hosted Synthetic Monitoring

For minimal budget—GitHub Actions Cron:

# .github/workflows/synthetic-monitoring.yml
name: Synthetic Monitoring

on:
  schedule:
    - cron: '*/5 * * * *'  # Every 5 minutes
  workflow_dispatch:

jobs:
  check-critical-paths:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Install Playwright
        run: npx playwright install chromium

      - name: Run synthetic checks
        env:
          APP_URL: ${{ vars.PRODUCTION_URL }}
          TEST_USER_EMAIL: ${{ secrets.SYNTHETIC_USER_EMAIL }}
          TEST_USER_PASSWORD: ${{ secrets.SYNTHETIC_USER_PASSWORD }}
        run: npx playwright test tests/synthetic/

      - name: Notify on failure
        if: failure()
        run: |
          curl -X POST "$SLACK_WEBHOOK" \
            -d '{"text": "Synthetic monitoring FAILED: checkout flow"}'
        env:
          SLACK_WEBHOOK: ${{ secrets.SLACK_WEBHOOK }}

Limitation: GitHub Actions Cron doesn't guarantee exact execution time—delays up to 5-15 minutes. For precise 1-minute monitoring, use managed service.

Managing Test Data

Synthetic tests require test accounts and data:

• Separate [email protected] user with production account
• Payment method: test Stripe card (4242 4242 4242 4242)
• Mark orders from synthetic user with tag, exclude from business analytics
• Regularly clean test data (cart, drafts)

Metrics and Alerts

Key synthetic monitoring metrics:

• Availability %—percentage of successful runs
• Step duration—time per individual step
• Total flow duration—total scenario time
• First failure step—where it broke

Alert: if 2 of 3 checks from different regions fail → critical alert to PagerDuty.

Timeline

• Checkly / Datadog Synthetic (managed)—2-3 days for scenarios
• GitHub Actions cron + Playwright—2-3 days
• Test accounts + data cleanup—1 day
• Alerts + incident management integration—1 day