Brian Devins

Good catch, x-b3-span ID in our implementation is used as the parent ID for the server span

We don't support the zipkin shared spans concept, in zipkin the "transport" span is shared by the client and server, which is why spanid and parent are needed so that...

You can configure sleuth the have the same behavior by using `spring.sleuth.supports-join=false` in case you need to run in a mixed environment

Ah, OK. I see now what you are seeing. The reason there aren't headers in the webflux request object is that we add them in the netty instrumentation. If you...

Span ID on client side becomes Parent Span ID on server side. Each webclient call will get it's own Span ID so the MDC will be different. As for logging...

Ok, that does sound like a bug. If you can create a reproducer then we can look at making a failing test to work against

@sirinn thank you for the input, to help us scope the effort could you tell us what vertx modules you are using? IE: for vertx 3 we support web, rx,...

I don't know when we will end up being able to upgrade to vert.x 4, to get something working you can use the Opentracing support that Vertx provides: https://vertx.io/docs/vertx-opentracing/java/ Then...

Sorry folks for the delay, to answer a couple questions: Yes you can disable automatic instrumentation using system properties, `dd.trace.integration..enabled=false`. So for vert.x I would expect probably wanting to disable...

I haven't taken the time to review this just yet, but just so you know, the current way we interpret the JAX-RS routes is fundamentally wrong so you may not...