Page whoever is on call when a check fails. WatchCron triggers a PagerDuty incident and resolves it when the check recovers, with one dedup key per monitor.
How it works#
WatchCron sends events through PagerDuty's Events API v2. When a monitor fails, WatchCron triggers an incident on the service you configure. When the monitor recovers, WatchCron sends a resolve event with the same dedup key, and PagerDuty closes the incident automatically. No manual cleanup.
One dedup key per monitor means a flapping check doesn't open ten incidents. Repeat failures update the existing one until recovery comes through.
Before you start#
You need:
- A PagerDuty account with permission to add integrations to a service.
- A WatchCron account on any plan. PagerDuty alerts work on Free, Starter, Pro and Business.
If you don't have a PagerDuty service yet, create one first. Most teams use one service per environment (production, staging) or per team.
Step 1 — Get an integration key in PagerDuty#
Open Services → Service Directory and pick the service where incidents should land (or create a new one).
Go to the Integrations tab, click Add an integration, and choose Events API V2. After adding it, expand the integration row and copy the Integration Key. It's a 32-character hex string.
| Where | What to do |
|---|---|
| Services → Service Directory | Open (or create) the target service |
| Integrations tab | Add an integration → Events API V2 |
| Integration row | Expand it, copy the Integration Key |
Pick Events API V2, not V1. The older version doesn't support resolve events with a dedup key, so auto-recovery won't work.
Step 2 — Add the channel in WatchCron#
Go to Alerts → Channels → Add channel and choose PagerDuty.
| Field | What to enter |
|---|---|
| Label | Something recognisable, e.g. "Production" |
| Integration (routing) key | The 32-character key from PagerDuty |
Click Save.
Step 3 — Test and assign monitors#
Click Send test. PagerDuty should create a test incident on the service. Check the PagerDuty dashboard to confirm it arrived, then resolve it manually.
Open any monitor and tick the PagerDuty channel under Alert channels. From now on, failures trigger real incidents routed through your escalation policy.
What to paste where#
| WatchCron field | Value | Where you find it |
|---|---|---|
| Integration (routing) key | R0123456789abcdef0123456789abcdef | Service → Integrations → Events API V2 |
What WatchCron sends#
Two event types:
- Trigger: when a monitor goes down or misses a heartbeat. Includes the monitor name, reason, and a link to the WatchCron incident page.
- Resolve: when the monitor recovers. Uses the same dedup key, so PagerDuty closes the matching incident.
Your on-call rotation, escalation rules, and notification preferences all apply as usual. WatchCron just fires the event; PagerDuty handles who gets paged and how.