Federico Mon

> @gnufede Can you fold these variants into the new `set_http_meta` scenario? I can, but then we'll have to be very careful about all the parameters of the config to...

Hi @hdadkhah1, thank you for reporting this issue. Can you confirm that 2.7.1 doesn't show this behavior? Or what version are you upgrading from?

> I updated the lib separately, do you still want to set the other variables? They are for a new feature that is about to go beta, it would be...

I guess this was some CI hang, but could you double check, @brettlangdon ? > test_tracer_usage_multiprocess - test_rand.py 7m 1.49s (+6m 57.58s, +10696%) - [Details](https://app.datadoghq.com/ci/test/AgAAAZALWQA9x7iwSAAAAAAAAAAYAAAAAEFaQUxXUUhWQUFBQlM3eElSNDI3cUxTcgAAACQAAAAAMDE5MDBiNTktMmU0NS00NzJjLWIzYzYtNzMyOGZhMDk3Y2Nm?colorBy=service&spanViewType=metadata&env=none&spanID=11474191554179990409&eventStack=AgAAAZALWQA9x7iwSAAAAAAAAAAYAAAAAEFaQUxXUUhWQUFBQlM3eElSNDI3cUxTcgAAACQAAAAAMDE5MDBiNTktMmU0NS00NzJjLWIzYzYtNzMyOGZhMDk3Y2Nm&currentTab=performance)

Switched back to draft as it apparently breaks all iast standalone tests

Hello @ruslansuprun thank you for reporting this issue. Can you tell us what products or environment variables you have enabled? For example `DD_IAST_ENABLED` or `DD_PROFILING_ENABLED`

Thank you, This CPU spike, does it also happen if you disable `DD_LOGS_INJECTION` or `DD_PROFILING_ENABLED`? The spike does not happen in 2.8.0 with the same env variables?

Addressed by https://github.com/DataDog/dd-trace-py/pull/10472

> If the goal is to de-vendorize pybind11, the correct approach would be to add the dependency in [build-system] requires in the pyproject.toml and import it in CMake using find_package(pybind11...

> does this need to be backported? Just if we want to run the macrobenchmarks on 2.9, 2.10 branches and not only main