Connecting to telemetry
Component health
Live status and recent operationsActive incidents
0 activeTraffic and errors
RequestsErrors
Live traffic
Production and authorised synthetic QA traffic are shown separately.
Requests and errors over time
Components
Unknown means no reliable telemetry, not healthy.
Grouped errors
Equivalent failures are grouped with counts and representative traces.
| Error | Component | Occurrences | Affected | Rate/min | First / last seen | Deployment | Trace |
|---|
Clinical safety
Monitors approved Phase 3 safety invariants; it does not create new clinical rules.
| Invariant | Summary | Consultation ref | Detected | Version | Status |
|---|
Consultation traces
Search by consultation/reference ID, trace ID or QA test ID. Patient demographics are not required.
Enter a technical identifier to reconstruct the sampled operation timeline.
Deployments
Versions and build identities for before/after health comparison.
| Time | Component | Version | Cloudflare deployment | Source hash | Dataset | Active |
|---|
Performance
Latency percentiles are calculated from aggregated fixed latency histograms.
| Component | Requests | RPS | Success | Errors | p50 | p95 | p99 | Maximum |
|---|
Incidents
Acknowledging, investigating or resolving an incident preserves all underlying evidence.
| Severity | Incident | Component | Occurrences | Affected | First / last seen | Status | Action |
|---|
Admin audit
Administrator access, trace views and significant actions.
| Time | Admin ID | Action | Target | Outcome | Client |
|---|