David Elner

Rebased this branch on the latest 2.0 and added one more name change of `TRACER_VERSION` to `GEM_DATADOG_VERSION`. All changes should be complete, just need a passing build now. Handing off...

> Our application spends a considerable amount of time (p50 10ms, p95 25ms) adding tags to traces on every single request. @sircpl Can you clarify on this? It's really adding...

Generally speaking, the instrumentation is meant to (and built to) minimize overhead on the "critical path": the user-managed thread running operations for the user's application. To accomplish this, we try...

I'm interested in your use case, but I'm having trouble understanding the trace flow you're describing, and visualizing the desired output in the UI. Particularly, I'm unfamiliar with "firehose" and...

I see; this definitely helps a lot! If I'm understanding correctly, it's basically a "fan-in", where the trace of the process reading from the queue is downstream of 40 other...

I tried brainstorming what a visualization could look like (granted I'm not a designer)...  Does that roughly model what you're looking for? (For...

Just an update; our team is internally looking at ways to support this kind of behavior more broadly in tracing & APM. I'm sharing this example as a use case...

> I think this is only right if you are using the default instrumentation configuration. However, if you are using the [supported and documented feature to use a custom service...

Hey @bessey! Thanks for the report! We'll have someone take a look at this very soon, and get sense of what's going on. They may need to ask some more...

Hi @HeyNonster! Thanks for the report and the thorough explanation here. Rejecting objects seems reasonable, but the `Hash` may be essential in different contexts. This is my hesitation about making...