Trace Propagation
Learn how to connect events across applications/services.
If the overall application landscape that you want to observe with Sentry consists of more than just a single service or application, distributed tracing can add a lot of value.
In the context of tracing events across a distributed system, distributed tracing acts as a powerful debugging tool. Imagine your application as a vast network of interconnected parts. For example, your system might be spread across different servers or your application might split into different backend and frontend services, each potentially having their own technology stack.
When an error or performance issue occurs, it can be challenging to pinpoint the root cause due to the complexity of such a system. Distributed tracing helps you follow the path of an event as it travels through this intricate web, recording every step it takes. By examining these traces, you can reconstruct the sequence of events leading up to the event of interest, identify the specific components involved, and understand their interactions. This detailed visibility enables you to diagnose and resolve issues more effectively, ultimately improving the reliability and performance of your distributed system.
Here's an example showing a distributed trace in Sentry:
This distributed trace shows a Vue app's pageload
making a request to a Python backend, which then calls the /api
endpoint of a Ruby microservice.
What happens in the background is that Sentry uses reads and further propagates two HTTP headers between your applications:
sentry-trace
baggage
If you run any JavaScript applications in your distributed system, make sure that those two headers are added to your CORS allowlist and won't be blocked or stripped by your proxy servers, gateways, or firewalls.
Server-side SDKs handle traces automatically on a per-request basis. This means that SDKs will:
- Continue an existing trace if the incoming request contains a trace header.
- Start a new trace if the incoming request does not contain a trace header. This trace stays active until the response is sent.
If necessary, you can override the default trace duration by manually starting a new trace.
In version 8.0.0
and above of the Sentry Node SDK, distributed tracing will work out of the box.
Sentry.init({
dsn: "https://examplePublicKey@o0.ingest.sentry.io/0",
tracesSampleRate: 1,
});
If you don't want to use tracing, you can set up Custom Instrumentation for distributed tracing.
If you want to disable distributed tracing and ensure no Sentry trace headers are sent, you can configure your SDK like this:
Sentry.init({
dsn: "https://examplePublicKey@o0.ingest.sentry.io/0",
// Overwrite the defaults to ensure no trace headers are sent
tracePropagationTargets: [],
});
Remember that in order to propagate trace information through your whole distributed system, you have to use Sentry in all of the involved services and applications. Take a look at the respective SDK documentation to learn how distributed tracing can be enabled for each platform.
Our documentation is open source and available on GitHub. Your contributions are welcome, whether fixing a typo (drat!) or suggesting an update ("yeah, this would be better").