Milan van der Meer
Milan van der Meer
Any progress on this? Ive been building this PR manually for quite some time now and it would be great to get this merged to have a better CI/CD experience...
@chbatey Could someone from the akka-team have a look at this? Without this, AkkaManagement is completely broken on Fargate ECS. We are stuck on the 1.0.0-RC2 release because of it.
@chbatey No worries, I heard you were on a team event :) I made a PR in the meantime to fix this in an easier way. I think this PR...
I will have a look later today :) Looks interesting! You are indeed correct the current scenario is only tested with 1 transaction. Although I've tested this with a dummy-application...
@felixbarny I've spent some time investigating this issue again :) [Kamon's implementation](https://github.com/kamon-io/Kamon/blob/master/instrumentation/kamon-scala-future/src/main/scala-2.13/kamon/instrumentation/futures/scala/FutureChainingInstrumentation.scala) has some nice descriptions on how and why they instrument certain methods to make the Future context propagation...
@felixbarny Been debugging this a bit more again :) I have some strange behavior that I have a question around: The following code is instrumenting a specific method: ``` @Nullable...
@felixbarny Im not sure how the `getActive` and the `currentTransaction` can give a different result though. In the following test: ``` val fs = (1 to 10).map(transactionNumber => Future {...
@felixbarny It indeed looks like the spans are correctly passed across threads and being linked to the correct `Future`. Meaning that `tracer.getActive()` will give the expected results. As you mention,...
@alejandrohdezma Yea, it is indeed an edge case in case you want to rerun scala-steward from scratch ignoring manually closed PRs from the past. So an environment variable makes total...