Dynatrace export

You can now export Hookdeck Event Gateway metrics to Dynatrace, alongside existing Datadog, New Relic and Prometheus support. The integration uses the Dynatrace Metrics API to send request, event and delivery attempt metrics directly to your Dynatrace SaaS environment.

The following metrics are exported:

• Requests (total, accepted, rejected, ignored)
• Events (total, successful, failed)
• Delivery attempts (total, delivered, failed)
• Response latency (average, p95, p99)
• Queue depth per destination

Each metric includes project and organization dimensions, plus the name of its source, connection or destination, so you can filter and alert on them in Dynatrace.

Response latency percentiles are also new for the other exports. Datadog and New Relic now receive hookdeck.attempts.response_latency_p95 and _p99. Prometheus exposes hookdeck_response_latency_p95_ms and _p99_ms, and hookdeck_response_latency_ms now reports the average over the last 5 minutes instead of an all-time average.

Datadog and New Relic now receive queue depth as hookdeck.queue_depth.message_count. Latency and queue depth are sent to both as gauges instead of counts, so Datadog's default rollup no longer adds up per-second latency averages.

If your Datadog account already has hookdeck.attempts.response_latency, it's registered as a count. Change its type to gauge in the Datadog Metrics Summary. Graphs that already use rollup(avg) keep working.

Configure the integration in your project integrations by providing a Dynatrace access token with the Ingest metrics scope and your environment URL. Metrics will start appearing in Dynatrace within a few minutes.

Metrics export is available on the Growth and Enterprise plans.