01
Service health overview
Compare request throughput, error rate, p95 and p99 latency, health state, and recent versions across services.
Product · Application Performance Monitoring
Understand application health before the next incident.
Turn OpenTelemetry traces into service-level throughput, error, and latency views. Compare transactions, dependencies, errors, and deployments, then open the representative request without rebuilding context.
- service health model
- RED
- tail-latency context
- p50→p99
- APM investigation views
- 7
- native trace source
- OTLP
From fleet health to the request that explains it.
APM summaries and trace evidence stay on the same tenant, time, and storage plane.
02
Transactions and error groups
Rank operations by volume, failure, and latency, then inspect representative traces and sanitized backend errors.
03
Dependencies and topology
See service, database, messaging, and external HTTP dependencies with their impact on the critical path.
04
Deployment context
Compare stable and candidate releases so a regression can be tied to the version that introduced it.
workflow
A service-first path to root cause.
Start with behavior across the fleet, then narrow to an operation, request, and dependency.
- 101 · Survey
Find the service consuming the error or latency budget
Use RED metrics and health state to separate an isolated regression from a wider platform event.
- 202 · Segment
Compare transactions, versions, and environments
Narrow the change to a route, release, namespace, or production environment.
- 303 · Inspect
Open a representative trace or error
Read the span path and dependency timing behind the aggregate without leaving the investigation.
- 404 · Correlate
Confirm with logs, metrics, and profiles
Carry service, trace, and time context into the next signal instead of starting a new search.
proof / shipped
APM is derived from the traces you already own.
Service summaries, transactions, dependencies, errors, deployments, and trace drill-downs are shipped product surfaces.
- Service throughput, error rate, p95, and p99
- Transaction and dependency ranking
- Grouped backend errors with representative samples
- Release-aware service analysis
Product · Application Performance Monitoring
Understand application health before the next incident.
Turn OpenTelemetry traces into service-level throughput, error, and latency views. Compare transactions, dependencies, errors, and deployments, then open the representative request without rebuilding context.