Find and query your Linkerd data

After setting up the OpenTelemetry Collector for Linkerd, your mesh telemetry flows into New Relic as Metric, Span, and Log events scoped under your Linkerd entity.

Find your Linkerd entity

Go to one.newrelic.com and select Entity Explorer.
Search for your cluster name.
Select your Linkerd entity to open the built-in dashboard.

The built-in dashboard covers request rate, latency p50/p95/p99, success rate, TCP connections, mTLS certificate status, control plane health, and meshed pod inventory.

Linkerd dashboard

The Linkerd entity dashboard provides an overview of your service mesh telemetry. It includes:

Section	Panels
Top-line	TCP Connections, Requests Served by Proxies
Traffic	Inbound Rate, Outbound Rate
Health	Success Rate, Error Rate, HTTP Status Codes
Latency	Request Latency p50/p95/p99, Outbound Latency p99
Security	mTLS Certificate Expiry, Cert Refresh Rate
Control Plane	Authorization Policy, Queue Depth, CP Request Rate
Topology	Proxy Memory Usage, Operations by Service
Inventory	Meshed Pods

Example NRQL queries

Replace <YOUR_CLUSTER> with your actual cluster name in each query.

Purpose	NRQL query
Request rate per service	`SELECT rate(sum(request_total), 1 minute) AS 'Req/min' FROM Metric WHERE k8s.cluster.name = '<YOUR_CLUSTER>' AND linkerd_control_plane_component IS NULL FACET k8s.deployment.name SINCE 30 minutes ago TIMESERIES`
Latency p99 per service	`SELECT histogram(response_latency_ms, width: 50, slide_by: 10) FROM Metric WHERE k8s.cluster.name = '<YOUR_CLUSTER>' AND linkerd_control_plane_component IS NULL FACET k8s.deployment.name SINCE 30 minutes ago`
HTTP success rate	`SELECT sum(response_total) AS 'Total responses', filter(sum(response_total), WHERE classification = 'success') / sum(response_total) * 100 AS 'Success %' FROM Metric WHERE k8s.cluster.name = '<YOUR_CLUSTER>' FACET k8s.deployment.name SINCE 30 minutes ago`
mTLS certificate expiry	`SELECT latest(identity_cert_expiration_timestamp_seconds) AS 'Cert expires (epoch)' FROM Metric WHERE k8s.cluster.name = '<YOUR_CLUSTER>' FACET k8s.deployment.name SINCE 10 minutes ago`
Control plane queue depth	`SELECT latest(control_plane_queue_length) AS 'Queue depth' FROM Metric WHERE k8s.cluster.name = '<YOUR_CLUSTER>' AND linkerd_control_plane_component IS NOT NULL FACET linkerd_control_plane_component SINCE 10 minutes ago`
TCP open connections	`SELECT rate(sum(tcp_open_connections), 1 second) AS 'TCP connections' FROM Metric WHERE k8s.cluster.name = '<YOUR_CLUSTER>' FACET k8s.deployment.name SINCE 30 minutes ago TIMESERIES`
Distributed traces per service (APM path only)	`SELECT count(*) FROM Span WHERE k8s.cluster.name = '<YOUR_CLUSTER>' FACET service.name SINCE 30 minutes ago`
Verify proxy spans are being collected with Linkerd attributes	`SELECT count(*) FROM Span WHERE k8s.cluster.name = '<YOUR_CLUSTER>' AND linkerd_control_plane_component IS NOT NULL FACET service.name SINCE 10 minutes ago`

If the "Verify proxy spans" query returns 0, the transform/linkerd_component_inject processor is not running or linkerd_control_plane_ns is missing from spans. See Troubleshoot Linkerd monitoring for remediation steps.

Troubleshoot Linkerd monitoring

Fix common issues with metrics, traces, and logs not appearing in New Relic.

Collect Linkerd proxy logs

Optionally collect linkerd-proxy sidecar container logs.

Metrics reference

Full list of Linkerd metrics and resource attributes collected by the OTel Collector.