TriggersAPI
Worker healthyAPI docs

Dashboard

Live view of events flowing through TriggersAPI, refreshed every couple seconds.

Event journey
One event, fanned out to every consumer it reached.

No event selected yet.

Pipeline
Live flow of events from ingestion to delivery. A traveling dot means a stage's count just grew.
…Ingested
…Queued
…Delivering
…Delivered
…Dead
…Canceled

In flight (0)

Demo controls
Drive the pipeline live: break the receiver, stall the worker, or flood it with events.
Receiver mode

Controls how the demo webhook receiver responds, so you can watch retries and dead-lettering happen live.

Worker

Simulate crash abandons the next delivery mid-attempt. Its 60-second lease expires and another worker rescues it.

Load

Creates that many events at once, each fanned out to active endpoints.

Duplicate delivery, proved
At-least-once means a receiver can genuinely see the same event twice. Rather than claim that, this makes it happen: the worker sends successfully, dies before it can write down that it succeeded, and the delivery goes out again when its lease expires.

The demo receiver hasn't been handed anything yet. Run the reliability demo, or press the button above.

Two rows sharing one delivery id, one processed and one skipped, is the whole argument. The receiver keeps track of the delivery ids it has handled, which is exactly what every webhook provider asks of you and exactly why they all send a delivery id header.

Events received
…
Delivered
…
In queue
…
Dead
…
Queue depth
…
Delivery latency
…
Throughput
Ingested, delivered, and failed attempts per 5s bucket, last 2 minutes.
IngestedDeliveredFailed
No data yet
Event feed
Latest events, acked or not. Click a row for delivery detail.