Visualizes how services call each other, so you can understand dependencies and see where latency/errors originate across the call graph.
Get here: Service Overview → Topology tab (service‑scoped map); a per‑trace map is also available on the Trace Detail → Service Maps tab.
What it shows
- Nodes = services (and external/downstream targets).
- Edges = calls between them, annotated with latency / throughput (and error state).
- The service‑scoped map centers on the selected service and shows callers and callees; the per‑trace map shows only the services involved in that one request.
Data source: sampled spans (topology derived from parent/child relationships).
How to use it
- Start from the service’s Topology to see its dependencies at a glance.
- Follow the edge with the worst latency/error rate to the offending downstream service.
- Open a representative trace → Service Maps tab, and cross‑check the Latency tab to see which downstream service contributed most to a specific request.