Detailed steps to reproduce the problem?

If you call `insert_many()` from two different threads that are being run concurrently, the ObjectId generation will regenerate the random number used for the id each time the threads swap.  This ends up spending a lot of cpu time generating random numbers rather than actually inserting documents.

Definition of done: what must be done to consider the task complete?

Swapping threads wouldn't slow down compared to not swapping threads.  That is, using a thread pool shouldn't make inserts substantially slower.

The exact Python version used, with patch level:

$ python -c "import sys; print(sys.version)"

3.11.9 (main, Aug  6 2024, 13:17:01) [GCC 13.2.0]

The exact version of PyMongo used, with patch level:

$ python -c "import pymongo; print(pymongo.version); print(pymongo.has_c())"

4.8.0
True

Describe how MongoDB is set up. Local vs Hosted, version, topology, load balanced, etc.

docker of mongo:latest

The operating system and version (e.g. Windows 7, OSX 10.8, ...)

Ubuntu 24.04

Web framework or asynchronous network library used, if any, with version (e.g. Django 1.7, mod_wsgi 4.3.0, gevent 1.0.1, Tornado 4.0.2, ...)

Security Vulnerabilities

If you’ve identified a security vulnerability in a driver or any other MongoDB project, please report it according to the instructions here