Service Map
A live topology graph showing every service in your workspace, call relationships, and health status at a glance.
Service Map
The service map is a live topology graph showing every service in your workspace and the call relationships between them. It is available at /services in the dashboard.
What you see
- Instrumented services: Solid nodes with full metrics on hover (latency, error rate, throughput)
- Ghost services: Dashed-border nodes with caller-side metrics and an amber accent. These are dependencies detected from outbound calls but not yet instrumented. See Ghost Services.
- Edges: Directed arrows showing service-to-service call patterns, colored by health:
- Green: healthy — error rate and latency within baseline
- Amber: degraded — elevated error rate or latency
- Red: failing — significant error rate or latency anomaly
How the map is built
The service map is constructed from two data sources:
- Service registry: Every instrumented service registers itself when it first sends CEs. Ghost services are registered when they are first observed as outbound call targets.
- Call edges: Directed edges between services, aggregated from outbound CEs. Each edge tracks call count, error count, and latency percentiles (p50, p95).
The map updates incrementally as new CEs arrive — there is no batch refresh.
Coverage scorecard
The coverage scorecard is displayed alongside the service map. It shows the ratio of instrumented to total services as a progress bar, with ghost services ranked by call volume. The highest-impact instrumentation targets appear first.
Using the map
Click an instrumented service to see its detail panel: recent traces, anomalies, and outbound call patterns.
Click a ghost service to see caller-side metrics and the "Instrument this service" action, which opens setup instructions pre-filled with your workspace API key and the ghost service name.
Service metadata
Each service can have metadata attached: owner team, on-call contact, runbook URL, and Slack channel. This metadata appears in the service detail panel and in incident traces, so you can quickly identify who to contact during an incident.
Edit metadata from the service map by clicking the service and opening the metadata panel.
Relationship to the trace viewer
The service map shows the topology at the workspace level — all services and their call relationships over time. The trace viewer shows a single incident's causal chain. Ghost services appear in both views, but the service map gives you the system-wide picture while the trace viewer shows what happened during a specific incident.