Performance Monitoring

The monitoring module collects runtime telemetry, assesses latency thresholds, and triggers alert notifications when health targets are breached. It powers the built-in terminal performance dashboard and maintains rolling statistical windows of request durations and status codes. It provides operations teams with immediate visibility into gateway health.

Exports

Symbol	Kind	Signature	Purpose
AlertSeverity	type	type AlertSeverity = 'info' | 'warning' | 'critical'	Alert severity levels
Metrics	interface	interface Metrics	Metrics for alert evaluation
AlertRule	interface	interface AlertRule	Alert rule definition
Alert	interface	interface Alert	Active alert
AlertConfig	interface	interface AlertConfig	Alert configuration
PerformanceAlerter	class	class PerformanceAlerter	Performance Alerter
TimeRange	interface	interface TimeRange	Time range for historical queries
HistoricalMetrics	interface	interface HistoricalMetrics	Historical metrics data
MetricPoint	interface	interface MetricPoint	Single metric point
MetricAggregates	interface	interface MetricAggregates	Metric aggregates
WorkerMetrics	interface	interface WorkerMetrics	Worker metrics
RouteMetrics	interface	interface RouteMetrics	Route metrics
UpstreamMetrics	interface	interface UpstreamMetrics	Upstream metrics
PerformanceAlert	interface	interface PerformanceAlert	Performance alert
EventStream	class	class EventStream extends EventEmitter	Event stream for SSE
PerformanceDashboard	class	class PerformanceDashboard	Performance Dashboard

Notes

PerformanceDashboard

Renders live ANSI terminal charts displaying P50, P95, and P99 latency alongside error rates.

PerformanceAlerts

Configurable alerting engine triggering webhooks or logs when error rates or latency spike above limits.

Source

• src/monitoring/performance-alerts.ts
• src/monitoring/performance-dashboard.ts

See also

• Deep Dive: Observability