Roch Devost

Can you share the profile and the code that has the added overhead? If it's not something you can share publicly, I would recommend opening a [support](https://www.datadoghq.com/support/) ticket.

> If I understand the PR correctly the dependencies are sent by group of synchronous requires right ? So I'm assuming we can't wait for the deps to be loaded...

Is this something you can reproduce with a small snippet that you could share? Webpack configurations can greatly differ and there are many things that could go wrong, but it's...

@dandean We've recently found out that there was a bug in the `graphql` plugin that caused `graphql.resolve` spans to have a duration of 0. While they would still show up...

@dandean Were you able to validate if the issue is fixed with the latest version of the tracer and/or replicate with a snippet you can share?

@schmod We're going to be working on Apollo support soon, but the main issue we've been having is that we are lacking knowledge of how it's used in the wild....

> While we can't release our source-code, I can throw together a basic example if you'd like. However, I have a feeling that it will look extremely similar to a...

Is this still an issue to anyone in this thread? If so, we'd still need more information about the expected behaviour in order to fix the issue. If not, please...

@joshribakoff-sm This is actually the expected behaviour for arrays as we combine all items as a single span to avoid generating too many resolve spans, and also because it's unlikely...

Is the tracer initialized before any instrumented module is imported? This is a requirement for dd-trace to work properly and the behaviour is undefined and can change from version to...