Chapter 02 / 06

Trace Explorer

Following a slow request end-to-end

Step 1Read the total duration

The trace header shows the full end-to-end duration. Compare it to the baseline to see how much slower this request is than normal — that single number tells you the severity immediately.

Executive viewSee how engineers identify exactly where a slowdown is happening — and how quickly.

Where did the slowdown happen?

A request took 2.35s — well above the expected baseline. Tracefox traced the request across all 4 services and pinpointed the root cause in one step.

2.35s
Time in bottleneck span
shop-frontend
Responsible service
< 1 min
Time to root cause

DB connection pool saturation in payment-service causes checkout latency spike — 41% drop in conversion rate. Engineers saw this immediately from the trace — no log-hunting required.

Distributed trace

a1b2c3d4e5f67890abcdef1234567890SLOW2.35s
Bottleneck identified: shop-frontend / POST /checkout consumed 2.35s of the total 2.35s 100% of total trace time.
Service / Operation
0ms587ms1.17s1.76s2.35s

Click a span to inspect its attributes and events.

Span detail

shop-frontend
POST /checkout
SLOW
Duration
2.35s
Start offset
+0ms
Attributes
http.method:POST
http.url:/checkout
http.status_code:200
user.id:u-9182736
cart.item_count:3
cart.value_usd:127.5
Span ID
span-001