eBPF deployment using the Harness CD pipeline

You can deploy the Traceable eBPF agent in your Kubernetes environment using a Harness Continuous Delivery (CD) pipeline. This deployment uses the Traceable Helm chart and is integrated into your Harness project. The process involves:

• Creating a Helm connector to access the Traceable Helm repository.

• Storing the agent configuration in the Harness File Store.

• Creating a Harness service for the Traceable agent.

• Adding the service to a Harness deployment pipeline.

Once set up, this approach enables consistent, automated deployments of the Traceable eBPF agent across environments.

Before You Begin

Before starting the deployment, ensure that you have:

• Access to Harness CD with permission to create connectors, services, and pipelines.

• A running Harness Delegate in your environment.
  If a delegate is not available, you must install one during the connector setup.

• A valid Traceable token and the environment details provided by Traceable.

• Helm chart repository details for Traceable:

  • Helm Repo URL: https://helm.traceable.ai

  • Authentication: None (unless otherwise specified by your organization)

• Namespaces for instrumentation — know which Kubernetes namespaces require eBPF instrumentation.

Deployment Steps

This procedure has four main steps. First, you create a Helm connector in Harness to access the Traceable Helm repository. Next, you store the Traceable agent configuration in the Harness File Store. You then create a Harness service that references the Helm chart and configuration file. Finally, you add the service to a pipeline and deploy it.

Step 1: Create a Helm Connector

The Helm connector in Harness provides access to the Traceable Helm repository.

1. In your Harness project, go to Project Settings → Connectors.

2. Click New Connector and select HTTP Helm Repository.

3. Configure the connector:

   • Name: traceable-helm-connector

   • Helm Repo URL: https://helm.traceable.ai

   • Authentication: None

4. Select an existing delegate or install a new one.

5. Click Test Connection and then Finish.

For more details, refer to the Harness documentation: HTTP Helm Repo Connector Settings Reference.

Step 2: Create a File Configuration

You now create a YAML configuration file for the Traceable agent and store it in the Harness File Store.

1. In your Harness project, go to Project Settings > File Store.

2. Click Create New File and configure:

   • Name: traceable-agent-values

   • File Usage: Manifest

   • Contents: YAML configuration, for example:

token: <your-token>
environment: <your-env-name>
endpoint: api-<traceable-env>.traceable.ai
logLevel: info
daemonSetMirroringEnabled: true
ebpfCaptureEnabled: true
ebpfRunAsPrivileged: true
ebpfDeployOnMaster: true
daemonSetMirroring:
  matchSelectors:
    - field_selectors:
        - "metadata.namespace=<namespace that needs instrumentation>"

Replace <your-token>, <your-env-name>, and <namespace> with your actual values.

Step 3: Create a Service

The service in Harness defines how the Traceable agent is deployed.

1. Go to the Services section and click New Service.

2. Enter:

   • Name: traceable-agent

   • Service Definition Type: Inline

   • Deployment Type: Kubernetes

3. In the Manifests section, click Add Manifest and configure:

   • Manifest Type: Helm Chart

   • Specify Helm Chart Store: HTTP Helm

   • Helm Connector: Select traceable-helm-connector

   • Manifest Identifier: traceable

   • Chart Name: traceable-agent

4. Add the configuration file as an override:

   • Click Add Additional Override File

   • Select Values YAML and Harness

   • Manifest Identifier: config

   • File Path: Select the traceable-agent-values file created in Step 2.

Step 4: Add the Service to a Pipeline

Finally, integrate the Traceable agent service into a deployment pipeline.

1. Open an existing pipeline or create a new one.

2. Add a Deploy stage or edit an existing stage.

3. Select the traceable-agent service created in Step 3.

4. Configure the stage as required and save the pipeline.

For details, see Harness Pipeline Execution Walkthrough.