Metrics Help
Derived 1-minute telemetry signals, anomaly state, and signal drilldown.
Browse Signals
Filter by source, service, signal name, or date range. Each row shows a derived 1-minute metric signal with its latest anomaly state.
Anomaly State
Signals show OK, Warning, or Critical state based on active threshold and seasonal rules.
Drilldown
Click a signal row to open the Anomaly Details page with a chart of the metric over time and correlated telemetry links.
Filter Controls
| Control | Description |
|---|---|
| Search | Regex pattern matched against signal name and service. |
| Source | OTEL telemetry source (e.g. logs, traces, gauge). |
| Service | Multi-select dropdown. Filter to one or more services. |
| Signal | Filter to a specific signal name (e.g. error_rate, p99_latency). |
| Attr FP | Optional attribute fingerprint to narrow signals to a specific attribute combination. |
| Hours | Lookback window in hours (1–168). Overridden by explicit From/To values. |
| From / To | Explicit date-time range. When set, overrides the Hours lookback. |
Rules & Anomaly Detection
Anomaly state is driven by metric rules. There are two rule types:
- Threshold rules — trigger when a signal crosses a static warning or critical threshold.
- Seasonal rules — auto-generated rules that use historical per-bucket baselines to detect unusual patterns.
Manage rules on the Metric Rules page.
Ingest & API
| Endpoint | Description |
|---|---|
| POST /v1/metrics | OpenTelemetry OTLP JSON metric data points (gauge, sum, histogram). |
| GET /metrics |
UI page. Accepts q, source, service, signal, hours, from_ts, to_ts query parameters.
|