Expected Behavior

We would not expect more than 5% overhead from logging capabilities as industry standard. (Not sure if the problem is with the reactor project or micrometer capabilities)

Actual Behavior

Huge overhead from micrometer - more than 25-30%

Steps to Reproduce

Simple request in spring webflux (same jvm process) where we want to have trace-id and correlate logs. See the profiler results showing huge overhead in the context propagation for setting thread locals:

Possible Solution

Your Environment

• Reactor version(s) used:
  reactor-core:3.6.7
• Other relevant libraries versions (eg. netty, ...):
  spring-webflux:3.3.1
  io.micrometer:1.13.1
• JVM version (java -version): java17
• OS and version (eg uname -a): windows 11