Distributed tracing with Jaeger on an OVHcloud Managed Kubernetes Service

Find out how to set up distributed tracing on OVHcloud Managed Kubernetes Service with Jaeger.

Last updated 22nd March 2022


Jaeger is an open-source distributed tracing platform.


It can be used for monitoring microservices-based distributed systems:

  • Distributed context propagation
  • Distributed transaction monitoring
  • Root cause analysis
  • Service dependency analysis
  • Performance / latency optimization

Jaeger contains several components:


Read more about Jaeger architecture and components.

In this guide you will:

  • Install Jaeger Operator
  • Deploy Jaeger components
  • Access to the UI
  • Deploy your instrumented application
  • Visualize traces

You can use the Reset cluster function in the Public Cloud section of the OVHcloud Control Panel to reinitialize your cluster before following this tutorial.


This tutorial presupposes that you already have a working OVHcloud Managed Kubernetes cluster, and some basic knowledge of how to operate it. If you want to know more on those topics, please look at the OVHcloud Managed Kubernetes Service Quickstart.

You also need to have Helm installed on your workstation and your cluster. Please refer to the How to install Helm on OVHcloud Managed Kubernetes Service tutorial.


In this guide we will show you how to deploy Jaeger as distributed tracing platform backend and then how to deploy your instrumented application that will send their traces to Jaeger.

Installing Jaeger

For this tutorial we are using the Jaeger Helm chart.

Add the Jaeger Helm repository:

helm repo add jaegertracing https://jaegertracing.github.io/helm-charts
helm repo update

These commands will add the Jaeger Helm repository to your local Helm chart repository and update the installed chart repositories:

$ helm repo add jaegertracing https://jaegertracing.github.io/helm-charts
helm repo update
"jaegertracing" has been added to your repositories
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "sealed-secrets" chart repository
...Successfully got an update from the "prometheus-community" chart repository
Update Complete. ⎈Happy Helming!⎈

The Jaeger repository provides two charts: jaeger and jaeger-operator. For the guide, you will deploy the jaeger-operator chart, which makes it easy to configure a minimal installation.

To learn more about the Jaeger Operator for Kubernetes, consult the official documentation.

Install the latest version of Jaeger with helm install command:

helm install jaeger-operator jaegertracing/jaeger-operator --namespace observability --create-namespace --set rbac.clusterRole=true

This command will install the latest version of Jaeger operator and observability namespace:

$ helm install jaeger-operator jaegertracing/jaeger-operator --namespace observability --create-namespace --set rbac.clusterRole=true
manifest_sorter.go:192: info: skipping unknown hook: "crd-install"
NAME: jaeger-operator
LAST DEPLOYED: Thu Mar 17 12:06:49 2022
NAMESPACE: observability
STATUS: deployed
jaeger-operator is installed.

Check the jaeger-operator logs
  export POD=$(kubectl get pods -l app.kubernetes.io/instance=jaeger-operator -lapp.kubernetes.io/name=jaeger-operator --namespace observability --output name)
  kubectl logs $POD --namespace=observability

Thanks to the variable overriding rbac.clusterRole=true, you ask the operator to watch all namespaces.

Check the Jaeger Operator is running:

kubectl get pod -n observability
$ kubectl get pod -n observability
NAME                               READY   STATUS    RESTARTS   AGE
jaeger-operator-67f8dd68c9-5qj26   1/1     Running   0          3m5s

The simplest possible way to create a Jaeger instance is by creating a YAML file that will install the default AllInOne image. This “all-in-one” image includes: agent, collector, query, ingester and Jaeger UI in a single pod, using in-memory storage by default.

For this guide you will deploy Jaeger components through this simple way, which can be used for development, testing and demo purposes but for production strategy you can read the official documentation.

Once the jaeger-operator pod in the namespace observability is ready, create a jaeger.yaml file with the following content:

apiVersion: jaegertracing.io/v1
kind: Jaeger
  name: jaeger
    serviceType: LoadBalancer

In this YAML manifest file we specify that you want to access the Jaeger UI (jaeger-query) through a Load Balancer.

And apply it:

kubectl apply -f jaeger.yaml

These commands will create a new CRD Jaeger and an instance named jaeger:

$ kubectl apply -f jaeger.yaml
jaeger.jaegertracing.io/jaeger created

You can now check if the Jaeger instance is running with the following commands:

kubectl get jaeger
kubectl get pods -l app.kubernetes.io/instance=jaeger

Theses commands will check if the instances were created, list the Jaeger objects and list the pods that are running:

$ kubectl get jaeger
jaeger   Running   1.30.0    allinone   memory    4s

$ kubectl get pods -l app.kubernetes.io/instance=jaeger
NAME                        READY   STATUS    RESTARTS   AGE
jaeger-59ccc99bcc-zpscb   1/1     Running   0          80s

You can also check that all Jaeger services have been correctly deployed:

$ kubectl get svc -l app=jaeger
NAME                          TYPE           CLUSTER-IP     EXTERNAL-IP      PORT(S)                                  AGE
jaeger-agent                ClusterIP      None                      5775/UDP,5778/TCP,6831/UDP,6832/UDP      3d21h
jaeger-collector            ClusterIP               9411/TCP,14250/TCP,14267/TCP,14268/TCP   3d21h
jaeger-collector-headless   ClusterIP      None                      9411/TCP,14250/TCP,14267/TCP,14268/TCP   3d21h
jaeger-query                LoadBalancer   16686:30598/TCP,16685:30835/TCP          3d21h

Access to Jaeger UI

Now you can retrieve Jaeger UI URL with the following command:

export JAEGER_URL=$(kubectl get svc jaeger-query -o jsonpath='{.status.loadBalancer.ingress[].ip}')
echo Jaeger URL: http://$JAEGER_URL:16686

You should obtain the following result:

$ export JAEGER_URL=$(kubectl get svc jaeger-query -o jsonpath='{.status.loadBalancer.ingress[].ip}')

$ echo Jaeger URL: http://$JAEGER_URL:16686
Jaeger URL:

Open your browser and go to the Jaeger interface.

Jaeger Query

Deploy your instrumented application

In order to link your application to the Jaeger backend you need to use a tool like OpenTelemetry.


OpenTelemetry is a collection of tools, APIs, and SDKs. Useful to instrument, generate, collect, and export telemetry data (metrics, logs, and traces) to help you analyze your software’s performance and behavior.

OpenTelemetry integrates with popular libraries and frameworks such as Spring, Express, Quarkus, and with a lot of languages. Go to the documentation to see how to integrate your application.

For this guide you will deploy a Golang application, instrumented with OpenTelemetry, that will send traces to a provider: your Jaeger collector.

Our main.go file contains:

  • the import of the OpenTelemetry dependencies,
  • a tracerProvider method that initiates a connection to a Jaeger provider
  • a main() method that connects to the Jaeger collector you previously deployed and creates and sends a span each time the / HTTP route will be called
package main

import (

    tracesdk "go.opentelemetry.io/otel/sdk/trace"
    semconv "go.opentelemetry.io/otel/semconv/v1.7.0"

const (
    service     = "go-what-is-my-pod-with-tracing"
    environment = "development"
    id          = 1

func tracerProvider(url string) (*tracesdk.TracerProvider, error) {
    // Create the Jaeger exporter
    exp, err := jaeger.New(jaeger.WithCollectorEndpoint(jaeger.WithEndpoint(url)))
    if err != nil {
        return nil, err
    tp := tracesdk.NewTracerProvider(
        // Always be sure to batch in production.
        // Record information about this application in a Resource.
            attribute.String("environment", environment),
            attribute.Int64("ID", id),
    return tp, nil

func main() {

    // Tracer
    tp, err := tracerProvider("http://jaeger-collector-headless.default.svc.cluster.local:14268/api/traces")
    if err != nil {

    // Register our TracerProvider as the global so any imported
    // instrumentation in the future will default to using it.

    ctx, cancel := context.WithCancel(context.Background())
    defer cancel()

    // Cleanly shutdown and flush telemetry when the application exits.
    defer func(ctx context.Context) {
        // Do not make the application hang when it is shutdown.
        ctx, cancel = context.WithTimeout(ctx, time.Second*5)
        defer cancel()
        if err := tp.Shutdown(ctx); err != nil {

    tr := tp.Tracer("component-main")

    ctx, span := tr.Start(ctx, "hello")
    defer span.End()

    // HTTP Handlers
    helloHandler := func(w http.ResponseWriter, r *http.Request) {
        // Use the global TracerProvider
        tr := otel.Tracer("hello-handler")
        _, span := tr.Start(ctx, "hello")
        defer span.End()

        podName := os.Getenv("MY_POD_NAME")
        fmt.Fprintf(w, "Hello %q!", podName)

    otelHandler := otelhttp.NewHandler(http.HandlerFunc(helloHandler), "Hello")

    http.Handle("/", otelHandler)

    log.Println("Listening on localhost:8080")

    log.Fatal(http.ListenAndServe(":8080", nil))

The code source of the application is available on GitHub.

We already packaged a Golang application into a Docker image and pushed it in our ovhplatform Docker Hub repository so you can use it directly.

In order to deploy the application on your OVHcloud managed Kubernetes Service, create a deployment.yaml file with the following content:

apiVersion: apps/v1
kind: Deployment
  name: what-is-my-pod-with-tracing-deployment
    app: what-is-my-pod-with-tracing
  replicas: 3
      app: what-is-my-pod-with-tracing
        app: what-is-my-pod-with-tracing
      - name: what-is-my-pod-with-tracing
        image: ovhplatform/what-is-my-pod-with-tracing:1.0.2
        - containerPort: 8080
          - name: MY_POD_NAME
                fieldPath: metadata.name

This YAML deployment manifest file defines that our application, based on ovhplatform/what-is-my-pod-with-tracing:1.0.2 image will be deployed with 3 replicas (3 pods). We pass the pod name on environment variable in order to display it in our what-is-my-pod-with-tracing application.

Then, create a svc.yaml file with the following content to define our service (a service exposes a deployment):

apiVersion: v1
kind: Service
    app: what-is-my-pod-with-tracing
  name: what-is-my-pod-with-tracing
  - port: 8080
    app: what-is-my-pod-with-tracing
  type: LoadBalancer

Apply the deployment and service manifest files to your cluster with the following commands:

kubectl apply -f deployment.yaml
kubectl apply -f svc.yaml

Output should be like this:

$ kubectl apply -f deployment.yml
deployment.apps/what-is-my-pod-with-tracing-deployment created

$ kubectl apply -f svc.yml
service/what-is-my-pod-with-tracing created

You can verify if your application is running and service is created by running the following commands:

kubectl get pod -l app=what-is-my-pod-with-tracing
kubectl get svc -l app=what-is-my-pod-with-tracing

Output should be like this:

$ kubectl get pod -l app=what-is-my-pod-with-tracing
NAME                                                      READY   STATUS    RESTARTS   AGE
what-is-my-pod-with-tracing-deployment-84b56684d8-6kw6z   1/1     Running   0          3m
what-is-my-pod-with-tracing-deployment-84b56684d8-bcsxh   1/1     Running   0          3m
what-is-my-pod-with-tracing-deployment-84b56684d8-wbjmz   1/1     Running   0          3m

$ kubectl get svc -l app=what-is-my-pod-with-tracing
NAME                          TYPE           CLUSTER-IP    EXTERNAL-IP      PORT(S)          AGE
what-is-my-pod-with-tracing   LoadBalancer   8080:32365/TCP   3m

In order to generate traffic you need to get the external IP of your service:

export APP_URL=$(kubectl get svc what-is-my-pod-with-tracing -o jsonpath='{.status.loadBalancer.ingress[].ip}')
echo Application URL: http://$APP_URL:8080/

And then generate traffic with curl command:

curl http://$APP_URL:8080/

You should obtain the following result:

$ export APP_URL=$(kubectl get svc what-is-my-pod-with-tracing -o jsonpath='{.status.loadBalancer.ingress[].ip}')

$ echo $APP_URL

$ curl http://$APP_URL:8080/
Hello "what-is-my-pod-with-tracing-deployment-84b56684d8-6kw6z"!%

$ curl http://$APP_URL:8080/
Hello "what-is-my-pod-with-tracing-deployment-84b56684d8-6kw6z"!%

$ curl http://$APP_URL:8080/
Hello "what-is-my-pod-with-tracing-deployment-84b56684d8-wbjmz"!%

Visualize traces

Open your browser and go back to the Jaeger interface (http://$JAEGER_URL:16686).

Jaeger query services

You should now see two available services:

  • go-what-is-my-pod-with-tracing
  • jaeger-query

Select go-what-is-my-pod-with-tracing service and click on Find Traces button.

Jaeger query traces

You can now click in a trace and visualize useful information.

Jaeger query trace details


Delete Jaeger components (created by the operator):

kubectl delete -f jaeger.yaml

Wait until the components are deleted and then you can uninstall the operator.

To uninstall Jaeger Operator, as you installed it through Helm, you can use helm uninstall command in order to delete the Jaeger Helm installed chart:

helm uninstall jaeger-operator -n observability

Delete the observability namespace:

kubectl delete ns observability

And delete the deployed application:

kubectl delete -f deployment.yaml
kubectl delete -f svc.yaml

Go further

Join our community of users on https://community.ovh.com/en/.

Czy ten przewodnik był pomocny?

Zachęcamy do przesyłania sugestii, które pomogą nam ulepszyć naszą dokumentację.

Obrazy, zawartość, struktura - podziel się swoim pomysłem, my dołożymy wszelkich starań, aby wprowadzić ulepszenia.

Zgłoszenie przesłane za pomocą tego formularza nie zostanie obsłużone. Skorzystaj z formularza "Utwórz zgłoszenie" .

Dziękujemy. Twoja opinia jest dla nas bardzo cenna.

Inne przewodniki, które mogą Cię zainteresować...

OVHcloud Community

Dostęp do OVHcloud Community Przesyłaj pytania, zdobywaj informacje, publikuj treści i kontaktuj się z innymi użytkownikami OVHcloud Community.

Porozmawiaj ze społecznością OVHcloud

Zgodnie z Dyrektywą 2006/112/WE po zmianach, od dnia 1 stycznia 2015 r., ceny brutto mogą różnić się w zależności od kraju zameldowania klienta
(ceny brutto wyświetlane domyślnie zawierają stawkę podatku VAT na terenie Polski).