fix: resolve asyncio event loop bug when gevent is installed [backport 2.19]

[github-actions]

Backport b769e1e from #11904 to 2.19.

Description

Reverts a change (4e312788f9dc4a7c65e2e69c2094be3313a6f858) that introduced a regression in asyncio event loops when gevent is installed. This issue cannot be reproduced on macOS; it was detected on Ubuntu 24.

Background

• In ddtrace v2.11.0, the ddtrace-py introduced support for crash tracking, and the crash tracker is started via ddtrace-run or by importing ddtrace.auto (here).
• Before the crash tracker is started, it reads the agent URL using the ensure_binary function.
• The ensure_binary function imports unittest.mock, which imports asyncio as a side effect.
• After crash tracking is started, ddtrace unloads all modules that were imported during the setup of ddtrace features (here). This includes asyncio.
• At this point, asyncio has been added to and then removed from sys.modules.
• When asyncio is imported in a user's application event loops fail to be set. This is seen in the script below (the script was run on Ubuntu 24 with gevent>=24)

Script

import asyncio
import time

loop = asyncio.get_event_loop()
loop_id = id(loop)
new_loop = asyncio.new_event_loop()
new_loop_id = id(new_loop)

print(f"old: {loop_id} new: {new_loop_id}")
asyncio.set_event_loop(new_loop)

check = asyncio.get_event_loop()
check_id = id(check)
print(f"check: {id(check)}")


while check_id != new_loop_id:
    print("MISMATCH")
    print(f"check: {id(check)}")
    time.sleep(1)
    check = asyncio.get_event_loop()
    check_id = id(check)

Output

MISMATCH
check: 131621237174144
MISMATCH
check: 131621237174144
MISMATCH
check: 131621237174144
MISMATCH
check: 131621237174144
MISMATCH
check: 131621237174144
MISMATCH

Next steps

Investigate module unloading and asyncio incompatibility

Checklist

• PR author has checked that all the criteria below are met
• The PR description includes an overview of the change
• The PR description articulates the motivation for the change
• The change includes tests OR the PR description describes a testing strategy
• The PR description notes risks associated with the change, if any
• Newly-added code is easy to change
• The change follows the library release note guidelines
• The change includes or references documentation updates if necessary
• Backport labels are set (if applicable)

Reviewer Checklist

• Reviewer has checked that all the criteria below are met
• Title is accurate
• All changes are related to the pull request's stated goal
• Avoids breaking API changes
• Testing strategy adequately addresses listed risks
• Newly-added code is easy to change
• Release note makes sense to a user of the library
• If necessary, author has acknowledged and discussed the performance implications of this PR as reported in the benchmarks PR comment
• Backport labels are set in a manner that is consistent with the release branch maintenance policy

[github-actions]

CODEOWNERS have been resolved as:

releasenotes/notes/resolves-gevent-asyncio-incompatiblities-246028676b10bea9.yaml  @DataDog/apm-python
.github/workflows/test_frameworks.yml                                   @DataDog/python-guild @DataDog/apm-core-python
ddtrace/internal/compat.py                                              @DataDog/python-guild @DataDog/apm-core-python
tests/contrib/asyncio/test_lazyimport.py                                @DataDog/apm-core-python @DataDog/apm-idm-python

[pr-commenter]

Benchmarks

Benchmark execution time: 2025-01-14 18:51:43

Comparing candidate commit b007732 in PR branch backport-11904-to-2.19 with baseline commit 5824827 in branch 2.19.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 394 metrics, 2 unstable metrics.