Distributed Tracing

Updated View MarkdownOpen in ChatGPTOpen in Claude

OpenObserve distributed tracing follows each request as it moves through your microservices. This helps you find slow spots and fix problems in distributed systems.

Overview

With distributed tracing, you can see the full path of a request across many services. You can see how services depend on each other and find the root cause of slow requests. OpenObserve is built for cloud-native apps. It ingests, links, and analyzes traces at enterprise scale, and queries stay fast.

Traces page Distributed Tracing overview page

Key Features

Trace Ingestion

  • OpenTelemetry Native: Full support for the OpenTelemetry protocol. Traces are collected automatically.

OpenTelemetry Native

  • High Throughput: Ingester nodes scale out in HA mode to handle more spans as volume grows.

Trace Structure & Data

  • Hierarchical Spans: See full trace trees that show parent-child links between operations.

Hierarchical Spans

  • Rich Metadata: Capture span attributes, tags, and context in detail.

Rich Metadata

  • Custom Instrumentation: Use manual or automatic instrumentation.
  • Advanced Filtering: Search traces by service, operation, duration, errors, and custom attributes.

Trace Search

  • Performance Profiling: Find slow operations and bottlenecks across your distributed system.

  • Error Investigation: Quickly find and study failed requests and exceptions.

Error Investigation

Visualization & Insights

  • Waterfall Views: See how spans relate and how long each one takes, in an interactive view.

Traces waterfall view

  • Flame Graph, Trace Graph and DAG: Other views of a single trace. DAG appears for traces with LLM spans.

  • Spans: Search and list single spans across all traces.

  • Service Graph (Enterprise): See how your services call each other, with dependencies, throughput and error rates.

Service Graph in Graph View with edges between the shop services and a tooltip showing requests, errors and error rate for checkout

  • Service Catalog: See every service that sends traces, with latency, throughput and error rate.

Storage & Performance

  • Indexing: Choose an Index Type per field (for example Full text search or Secondary index) in the stream's Schema Settings.

Index Type in the stream Schema Settings

Integration & Standards

  • OpenTelemetry Ecosystem: Works fully with OpenTelemetry collectors and instrumentation libraries.
  • Jaeger and Zipkin Compatibility: Ingest traces from Jaeger and Zipkin through the OpenTelemetry Collector.
  • Kubernetes-Friendly: Runs in Kubernetes through the OpenTelemetry Collector. It can add service metadata to your traces.
  • API Access: Use REST APIs to read and analyze traces from code.
Was this page helpful?