Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.tallwatch.com/llms.txt

Use this file to discover all available pages before exploring further.

Tallwatch sends incident events to PagerDuty using the Events API v2. PagerDuty’s escalation policy then decides who gets paged — by app push, SMS, voice call, or all three. Setup takes about two minutes once you know which PagerDuty service should receive the alerts.

Before you start

  • A PagerDuty account with the Events API v2 integration available (any paid tier, plus the Free trial).
  • The PagerDuty service that should receive these alerts. A service in PagerDuty corresponds to a thing-that-can-fail. Pick or create the right one.

Step 1: Add the Events API v2 integration in PagerDuty

1

Open the service

In PagerDuty, navigate to Services → Service Directory, then click the service you want to receive Tallwatch alerts.
2

Add an integration

On the service page, click the Integrations tab, then Add an integration. Search for Events API V2 and select it.
3

Save the integration key

PagerDuty creates the integration and shows an Integration Key (a 32-character string). Copy it. You’ll paste it into Tallwatch next.
The Events API v2 integration is sometimes called “Routing Key” inside PagerDuty’s docs. They’re the same thing. Tallwatch’s form labels it Integration key.

Step 2: Add the channel in Tallwatch

1

Open the channels page

Go to Settings → Alerts → Channels in the dashboard and click Add channel.
2

Pick PagerDuty

Select the PagerDuty kind. Give the channel a name that matches the PagerDuty service, like PagerDuty — billing prod.
3

Paste the integration key

Paste the 32-character integration key into the Integration key field.
4

Save and test

Click Save, then click Send test alert. A test incident should open in PagerDuty within seconds, marked as triggered.
Delete the test incident in PagerDuty after the dry run, otherwise your on-call rotation gets paged for a test.

How the events map

Tallwatch translates incident lifecycle events to PagerDuty event actions:
Tallwatch eventPagerDuty event_actionWhat happens in PagerDuty
incident.openedtriggerNew incident, escalation policy fires
incident.resolvedresolveLinked PagerDuty incident is auto-resolved
Manual ack in Tallwatch(none)PagerDuty stays open; ack in PagerDuty instead
Manual resolve in TallwatchresolvePagerDuty incident is auto-resolved
The dedup_key is set to the Tallwatch incident’s UUID, so multiple trigger events for the same incident merge into one PagerDuty incident rather than spawning duplicates.

What the alert looks like

A PagerDuty incident with:
  • A summary like Tallwatch: Monitor "API healthcheck" is down (3 of 4 regions)
  • A severity of error for opened, no severity change on resolve
  • The Tallwatch incident URL in the Links section so on-call can deep-link back
  • Custom details listing the failing regions, the check error class, and the duration

Acknowledge and resolve from PagerDuty

PagerDuty is the only channel where the ack/resolve action syncs back to Tallwatch in v1.1 (planned). For v1, acknowledging or resolving the incident in PagerDuty silences PagerDuty’s pages but does not flip the Tallwatch incident state. Use the Tallwatch dashboard to acknowledge or manually resolve from our side.

Troubleshooting

Check Tallwatch’s dispatch history for the test. If PagerDuty returned a 4xx, the most likely cause is the integration key was revoked or pasted incorrectly. Regenerate the integration in PagerDuty and update Tallwatch.If PagerDuty returned 202 Accepted but no incident appears, the service may have a deduplication suppression in place — e.g. the same incident was recently resolved, and PagerDuty is suppressing the re-trigger. Wait the dedup window out, or trigger from a different monitor.
Open the dispatch row to see the error. Common causes:
  • invalid_event_action — Tallwatch is sending an event_action PagerDuty no longer recognises. File an issue with us.
  • Integration disabled — someone in PagerDuty disabled the integration. Re-enable it in the service’s Integrations tab.
The dedup_key must match between the trigger and the resolve. Tallwatch always uses the incident UUID as the dedup_key, so this should work automatically. If it doesn’t, the most likely cause is the Tallwatch incident was manually resolved in PagerDuty first — PagerDuty considers the dedup_key stale after manual resolve and a fresh trigger creates a new incident.
Yes. Add one Tallwatch channel per PagerDuty service, each with its own integration key. Bind them to different monitors or escalation policies.

Reference

PropertyValue
Channel kindpagerduty
Required configintegration_key (32-character Events API v2 routing key)
PagerDuty APIEvents API v2 (https://events.pagerduty.com/v2/enqueue)
Dedup keyTallwatch incident UUID
Severityerror
Send rate limit1 per 30 s per channel (queued, not dropped)
Retry policyUp to 3 attempts on 5xx, then queued for the next 30-s slot