SignalFx Distributed Tracing intends to be instrumentation agnostic; supporting & ingesting variety of popular open instrumentation libraries including OpenTracing, Zipkin & OpenCensus. So long as the tracer configured to send spans to SignalFx ingest endpoint uses the Zipkin v1/2 JSON wire format or Jaeger Thrift format, we will accept spans irrespective of how they were instrumented - via one of the above mentioned open libraries or a homegrown one. Our goal is to build on the shoulders of giant communities rather than re-invent the wheel with proprietary libraries/agents and give customers choice to decide what works for them without worrying about vendor lock-in.
For customers who have not yet instrumented their code, our default recommendation is to use OpenTracing to instrument since it has a growing ecosystem of library owners & frameworks instrumenting their code with it & use Jaeger tracer libraries to export spans to us - both of these are CNCF projects with a rapidly growing community behind them.
RPC layer(s) and service/web framework(s) are the best places to start when thinking about how to go about instrumentating your application - both of these will likely have a large coverage area and touch a significant number of transaction paths to give you baseline tracing coverage and visualize an end-to-end trace with a service-map.
Next you should identify services critical to your business and look for areas not covered by rpc or service/web frameworks. Within these services, identify high value transactions and critical paths - instrument enough of these.
- Guide to help strategize instrumentation for your distributed application
- RPC and Web Frameworks pre-instrumented with OpenTracing
- Service-Mesh is another increasingly popular way to quickly get rpc/inter-service visibility
This repository contains a set of simple example applications that demonstrate using SignalFx with various open source tracers. They are broken down by language/platform.
- SignalFx.Tracing: An example using the SignalFx.Tracing.OpenTracing NuGet Package to instrument .NET libraries and applications.
- Jaeger Java: our recommended tracer for Java.
- Zipkin Brave: Zipkin's instrumentation library for Java.
- Java Kafka tracing: an example of tracing Kafka using Zipkin Brave with OpenTracing bindings.
- Spring Boot Examples:
- Jaeger Spring Boot: An example of instrumenting a Spring Boot application with Jaeger
- Zipkin Brave Spring Boot: An example of instrumenting a Spring Boot application with Zipkin
- OpenCensus Spring Boot: An example using the OpenCensus project's Jaeger exporter for java.
- Jaeger Python: our recommended tracer for Python.
- OpenCensus Python: OpenCensus's instrumentation library with a Jaeger reporter in Python.
- Jaeger Go: our recommended tracer for Golang.
- OpenCensus Go: OpenCensus's instrumentation library with a Jaeger reporter in Golang.
- Jaeger Node.js
- Zipkin JS Tracer: Zipkin's instrumentation library for Javascript (currently supports a broader range of instrumentations than the Jaeger Node tracer and supports running in a browser).
- Jaeger Ruby: a Jaeger trace exporter for Ruby
- Istio
- Istio Mixer Adapter: We have an out of process adapter available for Istio. This is an example configuration for that adapter.
- Istio E-Commerce Application: An example application with Istio tracing using the SignalFx adapter and Envoy tracing.
- Envoy: the Envoy proxy can be configured to report trace spans to SignalFx.
- AWS App Mesh: An example E-Commerce application with deployment and configuration files for App Mesh on ECS.
- AWS Lambda: Examples for instrumenting spans for AWS Lambda written in Java, Python, Node, Go
For customers who have not instrumented their applications, or have done so in an OpenTracing-compatible fashion, we offer several SignalFx Tracing libraries. Their detailed documentation is available in their respective source locations:
Examples of those auto-instrumentation techniques are available in this repository: