Skip to main content

Datadog Alerts Integration

Overview​

Configure Datadog monitors to send alert notifications to OpsWorker for automatic investigation.

Prerequisites​

  • A Datadog account with monitor configuration access
  • OpsWorker cluster set up with the Kubernetes Agent installed
  • Your OpsWorker webhook URL (available in the portal under Integrations → Datadog)

Setup​

1. Get Your Webhook URL and Token​

  1. In the OpsWorker portal, go to Integrations
  2. Select your cluster
  3. Find Datadog, click Install, and copy the token the portal generates

All clusters POST to the same endpoint: https://alerts.ow-saas-prod-ext.opsworker.ai/alerts (dev: https://alerts-dev.opsworker.ai/alerts). The cluster token is not part of the URL. Datadog must send the token in a custom Authorization header instead.

2. Create a Webhook in Datadog​

  1. In Datadog, go to Integrations → Webhooks
  2. Click New Webhook
  3. Configure:
    • Name: opsworker
    • URL: https://alerts.ow-saas-prod-ext.opsworker.ai/alerts
    • Payload: leave as default (OpsWorker accepts the standard Datadog webhook format)
    • Custom headers: add Authorization: Basic <token>, using the token from the portal. This header is required. Without it the webhook is rejected.
  4. Click Save

3. Add Webhook to Monitors​

For each monitor you want OpsWorker to investigate:

  1. Edit the monitor
  2. In the Notify your team section, add: @webhook-opsworker
  3. Save the monitor

Alternatively, add @webhook-opsworker to a monitor template to apply it to multiple monitors at once.

4. Verify​

Trigger a test alert from one of your configured monitors. Check the OpsWorker portal under Alerts to verify signals are arriving.

Adding Query Capabilities​

The Datadog webhook covers alert ingestion only. There is no Datadog query or MCP integration. To let OpsWorker run metric and log queries during investigations, configure the Grafana MCP integration. Grafana can sit in front of Datadog datasources and expose them via PromQL for OpsWorker.

Supported Monitor Types​

OpsWorker can investigate alerts from these Datadog monitor types:

  • Metric monitors
  • APM monitors
  • Log monitors
  • Process monitors
  • Composite monitors

For best investigation quality, ensure monitors include Kubernetes context tags (kube_namespace, pod_name, deployment) so OpsWorker can map the alert to a resource.

Next Steps​