Kubernetes + Splunk OpenTelemetry Example

This guide demonstrates how to run Node.js and Python sample applications on Kubernetes (via Docker Desktop) with Splunk OpenTelemetry instrumentation.

Prerequisites

• Docker Desktop with Kubernetes enabled
• kubectl CLI installed and configured
• Splunk Observability Cloud account with:
  • Access Token (SPLUNK_ACCESS_TOKEN)
  • HEC Token (SPLUNK_HEC_TOKEN)

---

1. Start Kubernetes in Docker Desktop

1. Open Docker Desktop.
2. Go to Settings → Kubernetes.
3. Enable Kubernetes and wait until it's running.

---

2. Configure Splunk OpenTelemetry Collector

Edit splunk-otel-deploy.yaml and update the following environment variables:

env:
  - name: SPLUNK_ACCESS_TOKEN
    value: "<ADD-YOURS>"
  - name: SPLUNK_REALM
    value: "eu0"
  - name: SPLUNK_METRICS_ENABLED
    value: "true"
  - name: SPLUNK_TRACE_ENABLED
    value: "true"
  - name: SPLUNK_LOGS_ENABLED
    value: "false"
  - name: SPLUNK_HEC_TOKEN
    value: "<ADD-YOURS>"

Apply the configuration:

kubectl apply -f splunk-otel-deploy.yaml

---

3. Build and Deploy Applications

Build Docker images

# Build Node.js app image
cd nodejs-app
docker build -t apmtest-nodejs:latest .
cd ../
# Build Python app image
cd python-app
docker build -t apmtest-python:latest .
cd ../

Deploy to Kubernetes

# Deploy Node.js app
kubectl apply -f apmtest-nodejs-deploy.yaml

# Deploy Python app
kubectl apply -f apmtest-python-deploy.yaml

---

4. Access the Applications

• Node.js App: http://localhost:8000/
• Python App: http://localhost:8001/
• Do some test calls:

for i in {1..10000}; do curl http://localhost:8000/; curl http://localhost:8001/; sleep 1; done

---

5. Verify

Check if all pods are running:

kubectl get pods

Logs can be viewed with:

kubectl logs <pod-name>

---

6. Check signalfx AP

If traces are sent OK, you should find
apmtest-nodejs & apmtest-python
in
https://dev.signalfx.com/#/apm
filtering by Environment Dev