Two ways to summarize a single trace beyond the raw waterfall.
Latency tab (per‑service breakdown)
A pie + legend with Service · Duration · Percentage — each service’s merged busy time in the trace (overlapping same‑service spans merged, not double‑counted) as a % of total trace duration (sorted highest first; <0.01% for tiny shares). Colors here are assigned per service, independent of the waterfall’s category colors.
Service Maps tab (per‑trace map)
The call graph for this trace — nodes = services, edges = calls — so you can see the path and where a downstream dependency added time. See also Service Map and Topology.
AI Analysis / trace summary (when enabled)
When AI features are enabled, an AI Analysis tab and an AI trace summary provide a plain‑language summary of the trace and highlight likely problem areas.
How to use it
- Use Latency to confirm which service dominated total time.
- Use Service Maps to see whether a downstream dependency is the cause.
- Use the AI summary (if available) for a quick narrative and suggested focus.