metric
metric
Open the metric time-series explorer for a metric directly from the search bar.
Description
The metric command is a command-page: it paints the metric-explorer surface for the named metric — the same view as the Observability console's Metrics tab — sourced from OpenTelemetry metrics (otel_metrics).
Like asset, tree, and cloud, it is a page directive and short-circuits: no ClickHouse scan is issued and any search expression before it is ignored. The explorer fetches the metric's series directly.
Syntax
| metric <name> [service=<svc>]Arguments
name (required)
Syntax: metric <name>
Description: The metric to open (e.g. http.server.request.duration, system.cpu.utilization). Metric names commonly contain dots; quote a name only if it contains spaces.
service (optional)
Syntax: service=<svc>
Description: Open the explorer scoped to a single service (the otel_metrics.service_name column), so the chart shows only that service's series.
What the page shows
- A time-series chart of the metric over the selected range, with selectable aggregation (avg, sum, min/max, p50/p95/p99, rate)
- Group-by and tag filters to break the series down by dimension
- A jump-off to create a metric monitor from the current view
Examples
Open a metric
| metric http.server.request.durationA system metric
| metric system.cpu.utilizationScoped to one service
| metric http.server.request.duration service=checkoutUsage Notes
Terminal command: metric must be the final command in the pipeline.
No scan: the command does not read the logs table; it opens the curated explorer, which fetches its own data.
Service scoping: service=<svc> opens the explorer pre-filtered to that service (the promoted otel_metrics.service_name column). For other breakdowns, use the explorer's group-by / tag filters after it opens.
Requires OTLP data: populated from ingested OpenTelemetry metrics.