Jim Talbut You write:

The logstash encoder includes additional context data that I need to get into the MDC for the Json encoder (mainly distributed tracing span/trace IDs).

How is this data managed. Is it managed per thread?