[Python backend] Are messages logged in the model added to the current span of the OTel context?

[kelkarn]

Pretty much what the title asks for.

I noticed that the Python backend has a Logging section for how to use the pb_utils.Logger object.

The 'Triton Server Trace' Nvidia page lists how to set up OTel context propagation here. I am using Triton version 24.04, which is >= 24.01 meaning, my Triton server version supports context propagation from upstream callers.

My question is around the OTel span. If I use the pb_utils.Logger object to log messages from within a Python backend model on my Triton 24.04, will these log statements also be added to the span that is part of the current context/trace? I would like to have these Python backend model log messages grouped together with other log messages from the upstream caller/rest of the trace.