#Architecture

This architecture is built on the OpenTelemetry and Jaeger technology stack, achieving the full lifecycle management of distributed tracing. The system comprises five core modules: data collection, transmission, storage, querying, and visualization.

#Core Components

1. OpenTelemetry System

   • opentelemetry-operator
     A cluster-level Operator responsible for deploying and managing the otel-collector component, providing OTel automatic injection capability.
   • otel-collector
     Receives tracing data from applications, filters and batches it, and then forwards it to the jaeger-collector.
   • Tracing UI
     A self-developed visualization interface that integrates with the jaeger-query API, supporting multi-dimensional query conditions.

2. Jaeger System

   • jaeger-operator
     Deploys and manages the jaeger-collector and jaeger-query components.
   • jaeger-collector
     Receives tracing data forwarded and processed by the otel-collector, performs format conversion, and writes it to Elasticsearch.
   • jaeger-query
     Provides a tracing query API, supporting multi-condition retrieval including TraceID and labels.

3. Storage Layer

   • Elasticsearch
     A distributed storage engine that supports efficient writing and retrieval of massive Span data.

#Data Flow

• Writing Process
  Application -> otel-collector -> jaeger-collector -> Elasticsearch

  The application generates Span data via SDK or automatic injection, which is standardized by the otel-collector and subsequently persisted to Elasticsearch by the jaeger-collector.

• Query Process
  User -> Tracing UI -> jaeger-query -> Elasticsearch

  The user submits query conditions through the UI, and jaeger-query retrieves data from Elasticsearch; the UI visualizes the results based on the return data.