Linkerd Integration

Linkerd Integration in Dash0

Linkerd is an open-source service mesh designed for Kubernetes and cloud-native environments.
It provides security, observability, and reliability by seamlessly adding features like zero-config mutual TLS, dynamic request routing, and real-time telemetry.

This Dash0 integration enables users to visualize Linkerd’s telemetry data within their organization, providing insights into service health, traffic patterns, and latency without modifying their existing Linkerd deployment.

For more information, visit the Linkerd website.

Use Cases

Gain real-time visibility into your Linkerd-powered Kubernetes deployments with this integration:

Service Health Monitoring: Track error rates, durations, requests across your service mesh
Traffic Observability: Analyze service-to-service communication patterns and identify potential bottlenecks.
Performance Insights: Leverage real-time metrics to detect issues and optimize service reliability.

Overview

Linkerd provides built-in telemetry and observability features, capturing metrics for service-to-service communication, proxy behavior, and control plane operations. Starting with Linkerd 2.19, native distributed tracing support is built into the proxy, allowing you to collect traces without additional extensions. This integration enables you to forward Linkerd metrics and traces to Dash0 using the OpenTelemetry Collector.

For more details, refer to the Linkerd documentation:

Prerequisites

Before setting up Linkerd monitoring, ensure:

Linkerd is installed and running in your Kubernetes cluster
You have access to configure the OpenTelemetry Collector

You'll need an OpenTelemetry Collector deployed. Consider using:

Collecting Metrics

Collector Configuration

The OpenTelemetry Collector uses the Prometheus receiver to scrape Linkerd metrics. Add the following scrape configuration to your collector:

This configuration follows the official Linkerd Prometheus scrape settings and collects:

Linkerd control plane metrics (e.g., success rates, request volumes)
Linkerd proxy metrics (e.g., latencies, retries, errors)
Service mirror metrics (for multi-cluster setups)

Verification

Check collector logs for successful scrapes
Navigate to Metrics Explorer in Dash0
Search for metrics with prefix request_total, response_total, or route_*

Collecting Traces

Starting with Linkerd 2.19, the Linkerd-Jaeger extension is deprecated. Tracing can now be configured directly in the Linkerd control plane installation. Linkerd exports traces using the OpenTelemetry protocol, allowing you to forward trace data to any OpenTelemetry-compatible collector.

If you are currently using the Linkerd-Jaeger extension, refer to the migration guide for instructions on transitioning to the new approach.

Configuring Tracing in Linkerd

Using Helm

If you installed Linkerd using Helm, add the following values to your Linkerd control plane configuration:

Then upgrade your Linkerd installation:

Using the CLI

If you installed Linkerd using the CLI, use the upgrade command with tracing parameters:

The meshIdentity stanza is mandatory: Linkerd can only export traces to a collector within the mesh. Ensure your OpenTelemetry Collector namespace has the linkerd.io/inject: enabled annotation so that the collector pods are part of the Linkerd mesh.

Restarting Meshed Pods

After updating the Linkerd configuration, restart all meshed pods to apply the new tracing settings:

After completing these steps, you should start seeing Linkerd traces in Dash0.

For more details, refer to the Linkerd Distributed Tracing Documentation.

Technology

Linkerd

Overview