Service Map

The Service Map draws services as nodes and calls as directed edges. It helps identify dependencies, failure propagation, and high-latency connections without requiring you to know the topology in advance.

---

Explore the Map

• Node size and health styling summarize traffic and errors.
• Directed edges represent observed calls between services.
• Service-kind chips show or hide categories of nodes.
• Filter services searches the map by service name.
• Selecting a service isolates its immediate neighborhood; clear isolation to return to the full map.
• The map follows the selected environment, service, and time range.

---

Inspect a Service

Select a node to open its side panel.

The panel shows:

• Throughput, P99 latency, error rate, and span count.
• Throughput and error-rate trends.
• A link to Service Detail.
• A scoped link to Traces.
• A link to the service's logs in logX when logX is available.

---

Inspect a Connection

Select an edge to inspect the relationship between its caller and callee. The connection panel includes:

• Call volume.
• Average, minimum, and maximum latency.
• Error rate and last-seen time.
• The operation identifying the connection.
• Links to either service's detail and to traces from the caller.

Connection statistics come from observed trace relationships, so a missing edge generally means no matching relationship was recorded in the current scope.

---

Typical Workflow

1. An alert identifies a user-facing service with rising failures.
2. Open Service Map and isolate that service.
3. Inspect unhealthy incoming and outgoing connections.
4. Open the suspected dependency's Service Detail, traces, or logs.
5. Use browser Back to return to the previous map context.