Metrics docs

Where to find metrics

Open Metrics from the app navigation. The page summarizes delivery health for the selected time window, then breaks the data down by source, destination, and recent failures.

• Summary: total events, total deliveries, successful deliveries, failed deliveries, retry backlog, and open issues.
• Chart: event or delivery trends over time, grouped by the interval you choose.
• Sources: which vendor feeds or Dispatch apps are producing traffic.
• Destinations: which receivers are succeeding, failing, or slowing down.
• Recent failures: the newest failed attempts with response status and error summary.

Reading the summary

• Events counts webhook messages accepted by Relin.
• Deliveries counts attempts to send accepted events to configured destinations or recipient endpoints.
• Success rate helps show whether receivers are accepting traffic reliably.
• Failures point to handlers, URLs, signing secrets, or customer endpoints that need attention.
• Open issues connects metrics to the problems that need investigation.

Chart controls

Use the chart controls to switch between event volume, delivery volume, success rate, and latency views. Change the time bucket when a chart is too noisy or too compressed.

• Use a shorter window when debugging an active incident.
• Use a longer window when checking whether a destination has been degrading over time.
• Compare source and destination tables with the chart before replaying a broad set of events.

Common investigations

• Failure spike: open Recent failures, identify the affected destination, fix the receiver, then replay affected events.
• Traffic drop: check the source table and vendor setup. Confirm the vendor is still sending test or production events.
• Slow receiver: check destination latency, then add a rate limit or increase backoff while the receiver is repaired.
• Unexpected volume: filter destinations by event type or body value so handlers receive only the events they own.

What metrics are not

Metrics help you find where to investigate. The event detail page is still the best place to inspect one event's payload, inbound attempt, outbound attempts, and replay controls.