Understanding Smart Retries

SupaGuard uses Smart Retries to eliminate the most common pain point in synthetic monitoring: false alarms.

The Problem: Flaky Tests

In traditional monitoring, a "check" is binary: it passes or it fails. If a check fails, you get an alert immediately.

However, the internet is messy. A check might fail because:

• A DNS server in AWS us-east-1 glitched for 2 seconds.
• A specific CDN node timed out.
• A localized ISP issue occurred.

None of these are your fault, and often your site is working fine for everyone else. Yet, you wake up at 3 AM for a false positive.

The Solution: Multi-Region Verification

When SupaGuard detects a failure, we don't trust it immediately. Instead, we initiate a Smart Retry Protocol:

1. Initial Failure: A check fails in Location A (e.g., San Francisco).
2. Immediate Verification: We instantaneously spin up a new browser instance in Location B (e.g., London).
3. Confirmation:
   • If Location B passes, we mark the initial failure as a "Transient Network Issue" and do not alert you.
   • If Location B fails (and optionally Location C), we confirm the outage and trigger your alert policy.

Benefits

• Zero False Alarms: You only hear from us when real users are affected.
• Global Confidence: Confirms that an issue is widespread, not just a local glitch.
• No Manual Configuration: This logic is built-in and enabled by default for all checks.