Skip to main content

End to end setup

In this example, we're setting up an example project that:

  • has tracing enabled
  • exposes an endpoint to collect user data
  • creates and produces message into Kafka
  • processes messages 6 times while simulating a failure condition
  • simulates a succeeding condition at the end

Enable tracing across entire binary​

While KP produces spans, it has no knowledge on where the spans need to be sent and in which format. We expect libary user to set that up.

In this example we're setting up jaeger traces to send to tempo.

Start all dependencies that you need​

Here's a sample docker-compose.yaml that starts the dependencies that the demo requires.

Setup tracing such that trace-ids are exposed to the user​

Here's tracing setup that exposes traceparent header via response headers.

Produce a message with traces​

Producing a message is as simple as calling .Produce with a request context as a first parameters and the message as second parameter.

Configuring worker to fail multiple times​

Configure worker with retries, deadletters and tracing as shown

Execution​

Simply enter into v2/examples/full directory and simply run docker-compose up or docker compose up and wait for docker compose ps sleep command to indicate the container has exited (it'll take about 90 seconds)

Once docker compose logs web command indicates the server is listening, open the link in the browser and submit an email. Click on the link and refresh after a few seconds for traces to arrive.