Introduction

With Kubernetes 2.0 Agent, the Tracing Proxy is automatically installed when tracing is enabled for the cluster. This ensures seamless trace ingestion and management within the Kubernetes environment.

This document ensures a seamless setup process for users to enable trace ingestion with Kubernetes 2.0 agents.

Prerequisites

  1. Ensure that the Client-level Trace Management feature is enabled. For more details, refer to Enable Traces document.
  2. While installing the Kubernetes 2.0 Agent, enable the Enable Traces option and provide the required replica count.
    For more details, refer to the Installation document.
    Event Management

If tracing is enabled, Helm automatically installs tracing-proxy pods with the specified replica count.

OpsRamp Agents in Kubernetes

OpsRamp deploys two types of agents within the Kubernetes environment:

  • Master Agent: Runs as a Deployment (single replica).
  • Worker Agent Pod: Runs as a DaemonSet (one per node).

Both agents support trace ingestion, and users can choose the appropriate method based on their instrumentation setup.

Choosing the Right Trace Pipeline

  • Master Agent Trace Pipeline: Recommended when traces need to be exported through a centralized Kubernetes service.
  • Worker Agent Trace Pipeline: Recommended when the instrumentation solution exports traces on a per-node basis.

Trace Ingestion Configuration

Configure Trace Ingestion via Master Agent

To enable trace ingestion through the Master Agent, follow these steps:

  1. Instrument the application to send traces over gRPC or HTTP.
  2. Configure the following endpoints for trace ingestion:
    • gRPC Endpoint:
      opsramp-master-agent.opsramp-agent.svc.cluster.local:9081
    • HTTP Endpoint:
      opsramp-master-agent.opsramp-agent.svc.cluster.local:9082

If you want to modify the listen ports of the trace proxy, update the respective endpoint using the following Helm command:

--helm upgrade <release-name> <chart-name> --reuse-values --set traces.grpcPort=<port>  --set traces.httpPort=<port>

Configure Trace Ingestion via Worker Agent

To enable trace ingestion through the Worker Agent, follow these steps:

  1. Instrument the application to send traces over gRPC or HTTP.
  2. Configure the following endpoints for trace ingestion:
    • gRPC Endpoint:
      <Node_IP_OF_POD>:9081
    • HTTP Endpoint:
      <Node_IP_OF_POD>:9082

For detailed information on trace collection, refer to the Trace Collection document.