Alerts

Quorum-based alerting that only fires when an outage is real — no more 3 AM false alarms.

How quorum-based alerting works

PingGuard uses a quorum system to eliminate false positives. Instead of alerting on the first failed check, we require 2 of 3 regionsto confirm an outage before firing an alert. Here's the full evaluation flow:

  1. The alert evaluator runs every 60 seconds as a Vercel Cron job
  2. For each active monitor, it fetches the last 2 checks per region from the past 5 minutes
  3. A region is "confirmed down" only if both latest checks report failure
  4. The quorum threshold is ceil(total_regions x 0.66) — currently 2 of 3
  5. If the threshold is met and the monitor was previously up: mark as DOWN, create an incident, and fire a "Down" alert
  6. If the threshold is no longer met and the monitor was previously down: mark as UP, resolve the incident, and fire a "Recovery" alert

This means a brief network hiccup in a single region is silently absorbed. You only hear from PingGuard when something is genuinely wrong.

Setting up email alerts

Email alerts are available on all plans, including Free.

  1. Go to Settings > Alert Channels
  2. Click Add Channel and select Email
  3. Enter the email address to receive alerts
  4. Click the verification link sent to that address — alerts won't fire until verified

Alert emails include the monitor name, URL, failing regions, HTTP status code, and timestamp. Recovery emails also include the total downtime duration.

Setting up Slack alerts

Slack integration is available on Pro and Business plans.

  1. Go to Settings > Alert Channels
  2. Click Add Channel and select Slack
  3. You'll be redirected to Slack's OAuth flow — select your workspace and the channel where alerts should post
  4. After authorization, the channel appears as a verified alert destination

Slack messages use Block Kit formatting with color-coded status indicators, the failing URL, affected regions, and a direct link to the monitor detail page.

Alert deduplication and cooldown

PingGuard prevents alert fatigue with built-in deduplication:

15-minute cooldown

After a "Down" alert fires for a monitor, duplicate alerts for the same monitor are suppressed for 15 minutes — even if the evaluator continues to detect failure on each run.

Single recovery alert

When a monitor recovers, exactly one "Recovery" alert fires. The recovery email/Slack message includes the total downtime duration.

Flap detection

If a monitor goes down → up → down within 5 minutes, PingGuard consolidates this into a single "Intermittent" alert instead of sending separate down/recovery/down notifications.

This is especially useful for endpoints that experience transient issues — database connection pool exhaustion, occasional timeouts under load, or CDN cache misses that resolve themselves quickly.

Testing alerts

Each alert channel has a Test Alert button that sends a sample notification. Use this to:

  • Verify email delivery and check spam filters
  • Confirm Slack channel permissions and formatting
  • Validate webhook endpoint connectivity

Test alerts are clearly marked as test notifications and do not create incidents or affect monitor status.

Alert channels by plan

ChannelFreeProBusiness
Email✓✓✓
Slack—✓✓
Webhook——✓

Need more alert channels? See the Billing & Plans page for upgrade options.