feat!: Detect anomalous Datadog traces; upgrade to Python 3.11

[timmc-edx]

• Detect and log anomalous traces in DatadogDiagnosticMiddleware.
• Drop Python 3.8 support and change to Python 3.11 minimum due to testing needs.
• Add tests for existing middleware code

Manual testing:

1. Install in devstack, add middleware to lms/envs/devstack.py, and
   enable Waffle flag.
2. Make some requests and observe lack of new logging.
3. Add self.dd_tracer.current_root_span().finish() to start of
   process_view (faking an anomalous trace) and reinstall in devstack
4. Make some more requests and observe log output like the following:
   Anomalous Datadog local root span (duration already set): id = 66abf66b000000001168265a391d04a3; duration = 0.108 sec; worker age = 1.560 sec

Unit test improvements:

• Install ddtrace at the test.in level
• Add tests for new and existing middleware functionality
• Test with Python 3.11, not 3.8. This isn't just because we're upgrading
  to 3.11 across our repos, but also because I ran into a really weird bug
  with ddtrace and/or jsonschema that only occurs with Python 3.8:
  Installing ddtrace breaks tests with the jsonschema library DataDog/dd-trace-py#9950

---

Merge checklist:
Check off if complete or not applicable:

• Version bumped
• Changelog record added
• Documentation updated (not only docstrings)
• Fixup commits are squashed away
• Unit tests added/updated
• Manual testing instructions provided
• Noted any: Concerns, dependencies, migration issues, deadlines, tickets