Setting up the OpenTelemetry Demo in Kubernetes
The OpenTelemetry demo application can be deployed using Kubernetes.
Prerequisites
If you’d like to follow along, you’ll need the following software:
- git
- Docker
- A Kubernetes cluster v1.23+ (you can use minikube if running locally)
- Helm v3.9
- Splunk Observability Cloud
- Splunk Cloud Platform
You’ll want to have at least 4 GB of RAM available on your machine to run the demo application.
Create a new index
You can create a new index for the application logs. This is optional, but for our example, let’s create an index named “astronomyshop”.
- In Splunk Cloud Platform, navigate to Settings -> Indexes -> New Index.
- Enter the index properties you want, as shown in the following screenshot.
For further details on the index creation process, see Create events indexes.
Generate a Splunk HTTP Event Collector token
To use HTTP Event Collector (HEC) to send log data to Splunk Cloud Platform, we need to first generate a token.
- In Splunk Cloud Platform, navigate to Settings -> Data Inputs -> HTTP Event Collector -> New Token.
- Provide a name for the token, such as “astronomyshop_token”.
- Specify which index(es) you’d like the HEC token to send events to. In our case, we’ll send events to the “astronomyshop” index.
- After the token has been created successfully, make a note of its value for future reference.
For further details on the token generation process, see Create an Event Collector token on Splunk Cloud Platform.
Generate a Splunk Observability Cloud access token
While we’ll send log data to Splunk Cloud Platform, we’ll be sending metrics and traces to Splunk Observability Cloud. This requires an access token.
- In Splunk Observability Cloud, navigate to the Settings Wheel ☸ -> Access Tokens -> New Token.
- Enter a name for the token such as “AstronomyShop Token”.
- Because we'll use this token for both ingest and API, select the options shown in the following screenshot.
- Make a note of the token for future reference, as we’ll need it below.
For further details on the creation process, see Create an access token. The OpenTelemetry demo application generates approximately 5,000 custom metrics, primarily due to metrics related to Kafka and RPC. If you’d like to minimize custom metric usage within your Splunk Observability Cloud account, set a custom metric limit on the token used to instrument the demo application. For more information, see Manage data ingest using access tokens.
How to use Splunk software for this use case
Install the Splunk Distribution of the OpenTelemetry Collector for Kubernetes
Use the Splunk Observability Cloud guided installation for deploying the Splunk Distribution of the OpenTelemetry Collector for Kubernetes in our Kubernetes cluster. You can find the guidance at Data Management > Available integrations > Splunk OpenTelemetry Collector. Note the following settings:
- Enter the desired name for your Kubernetes cluster, such as “minikube”.
- If deploying the demo application to a public cloud environment such as AWS, Azure, or GCP, select the appropriate settings for Provider and Distribution.
Next, we’ll get a list of commands we can use to deploy the Splunk Distribution of the OpenTelemetry Collector to our Kubernetes cluster using Helm.
After these commands are executed, you should start to see data flowing into Splunk Observability Cloud after a minute or so.
Point to the Splunk Distribution of the OpenTelemetry Collector
Our demo application includes its own OpenTelemetry collector. But for our example, we want to leverage the Splunk Distribution of the OpenTelemetry Collector that’s already running in our Kubernetes cluster, which is already configured to export data to Splunk Observability Cloud.
To do this, we created a custom file named splunk/splunk-values.yaml and included the following environment variables:
env:
- name: OTEL_SERVICE_NAME
valueFrom:
fieldRef:
apiVersion: v1
fieldPath: "metadata.labels['app.kubernetes.io/component']"
# Ensure services report telemetry to the collector running on the same node
- name: OTEL_COLLECTOR_NAME
valueFrom:
fieldRef:
fieldPath: status.hostIP
- name: OTEL_EXPORTER_OTLP_METRICS_TEMPORALITY_PREFERENCE
value: cumulative
- name: OTEL_RESOURCE_ATTRIBUTES
value: service.version={{ .Chart.AppVersion }},deployment.environment=development
This tells each service to utilize the IP address of the node it runs on to connect to the collector.
We also updated the OTEL_RESOURCE_ATTRIBUTES for each service to include the deployment.environment.name attribute.
In our example, we set this to “development”, but feel free to change it if you wish.
Create a namespace for the OpenTelemetry Demo
Next, let’s create a separate namespace to deploy our demo application into.
kubectl create namespace otel-demo
Add the OpenTelemetry Helm Repository
Add the OpenTelemetry Helm repository using the following command:
helm repo add open-telemetry https://open-telemetry.github.io/opentelemetry-helm-charts
Then ensure it's up-to-date with:
helm repo update
Run the OpenTelemetry Demo
Run the demo application using the following command:
kubectl apply --namespace otel-demo -f ./splunk/opentelemetry-demo.yaml
After a minute or so you should see the application components appearing on the service map by navigating to APM -> Explore in Splunk Observability Cloud.
Next steps
The OpenTelemetry Demo is running, and is sending metrics and traces to Splunk Observability Cloud. Now, to get logs into Splunk Cloud Platform, let's look at Getting Kubernetes log data Into Splunk Cloud Platform with OpenTelemetry.