@road21 ^^
Why do you want to disable the finagle tracing mechanism instead of proving a different Tracer implementation?
@spockz I am using typed-schema (https://github.com/TinkoffCreditSystems/typed-schema) with finagle backend
And I can build context of my routes (reader context) only from Request class (see for example https://github.com/TinkoffCreditSystems/typed-schema/blob/master/modules/finagle-zio/src/main/scala/ru/tinkoff/tschema/finagle/routing/ZioRouting.scala)
I want to keep tracing headers in my context (for logging and propagating), that's why I need somehow to extract it from Request
And I can build context of my routes (reader context) only from Request class (see for example https://github.com/TinkoffCreditSystems/typed-schema/blob/master/modules/finagle-zio/src/main/scala/ru/tinkoff/tschema/finagle/routing/ZioRouting.scala)
I want to keep tracing headers in my context (for logging and propagating), that's why I need somehow to extract it from Request
You can read the current trace (according to finagle context) with something in Trace or Tracing. (Not near a pc right now so don’t have more detailed identifiers). So you don’t necessarily need to parse the B3 headers yourself
@mgedigian yes, you'd use a filter -- see: https://twitter.github.io/finatra/user-guide/thrift/filters.html. We have an access logging filter for Thrift: https://github.com/twitter/finatra/blob/develop/thrift/src/main/scala/com/twitter/finatra/thrift/filters/AccessLoggingFilter.scala. Note that logging the bytes is likely a security hole and we wouldn't recommend doing this but it's up to what you're trying to do. We have other tools and have been kicking around some ideas about how to capture requests for testing but it would be for capturing thrift requests your clients make to other services, not for requests to your thrift service.
when I kill the server in SBT console, and try to restart it, it is still bound to 0.0.0.0
(along with weird classloader issues throwing warnings)
yeah we were doing that for a while which was helpful for stuff like this, but there is some weirdness where the forked process doesn't inherit the SBT options of the parent (we noticed because it would run out of memory even though the parent was granted a lot). I should look into it again
thanks
[warn] Thread[Service Thread,9,system] loading com.twitter.jvm.Allocations$$anon$1$$anonfun$1 after test or run has completed. This is a likely resource leak
if anyone knows anything about this
I should look into pants one of these days
As a Buf or as a String
I know it sounds utterly stupid, not being able to find it