Troubleshoot the Helm deployment

Use these checks to investigate common issues with the Helm deployment.

Check pod status

BASH 
kubectl get pods -l app.kubernetes.io/name=splunk-opentelemetry-collector-for-kafka

View logs

BASH 
kubectl logs -l app.kubernetes.io/name=splunk-opentelemetry-collector-for-kafka -f

Check configuration

BASH 
# View the generated OpenTelemetry config
kubectl get configmap -l app.kubernetes.io/name=splunk-opentelemetry-collector-for-kafka -o yaml

Check collector health

BASH 
# Port forward to health endpoint
kubectl port-forward svc/<release-name>-splunk-opentelemetry-collector-for-kafka 13133:13133

# Check health
curl http://localhost:13133/

# Or connect directly to a pod:
kubectl port-forward -l app.kubernetes.io/name=splunk-opentelemetry-collector-for-kafka 13133:13133

Common issues

Pods not starting

  • Confirm that the Secrets exist and use the required keys.

  • Confirm that the Kafka brokers are reachable.

  • Check the resource limits.

  • Review pod events by running kubectl describe pod <pod-name>.

No data in Splunk

  • Confirm that the HTTP Event Collector (HEC) token is correct.

  • Confirm that the Splunk HEC endpoint is accessible.

  • Review collector logs for errors.

  • Confirm that the pipeline configuration uses the correct receiver and exporter names.

  • Check network connectivity from the cluster to the Splunk HEC endpoint.

Authentication failures

  • Confirm that the Secrets exist and include the password key for Kafka authentication.

  • Confirm that the Secret names match the configuration.

  • Check that the username and password are correct.

  • Review the Kafka broker authentication requirements.

Configuration errors

  • Confirm that the receiver and exporter names in the pipelines match their configured names.

  • Confirm that the configuration includes at least one receiver, exporter, and pipeline.

  • Review the generated ConfigMap for syntax errors.

  • Render the Helm templates by running helm template . -f values.yaml.