A slow request crosses API Gateway → Auth → Appointment → Payment → Notification. Render the span waterfall and spot the culprit instantly. Inject latency into any service and watch the waterfall shift to point at it.
One booking request flows through five services. Inject latency into any of them and watch the trace waterfall shift to point at the slow one.
Slow service
+7.70s
Trace waterfall · trace_id 7f3a91total 8.02s
API Gateway
30ms
Auth
50ms
Appointment
80ms
Payment
7.82s
Notification
40ms
time →
🧵 Slowest span: Payment at 7.82s — 98% of the whole request. The trace points right at it.
What just happened
▹A distributed trace follows one request across every service it touches. Each hop is a span with a start time, duration, and status — and a trace stitches them into one timeline.
▹Drawn as a waterfall, the slow service is impossible to miss: its bar is the long one, and every span after it is pushed to the right. You go from 'the request is slow' to 'it's the Payment service' in one glance.
▹This is why tracing is the observability superpower for microservices: without it you'd guess across five services; with it the timeline points straight at the bottleneck.