fix(serve): ensure a single event loop for python 3.11

[willkill07]

Description

This PR ensures that a single event loop is in use at all times when issuing nat serve

• catch KeyboardInterrupt cleanly to reduce error propagation
• manually start the server rather than call uvicorn.run()

Closes

By Submitting this PR I confirm:

• I am familiar with the Contributing Guidelines.
• We require that all contributors "sign-off" on their commits. This certifies that the contribution is your original work, or you have rights to submit it under the same license, or a compatible license.
  • Any contribution which contains commits that are not Signed-Off will not be accepted.
• When the PR is ready for review, new or existing tests cover these changes.
• When the PR is ready for review, the documentation is up to date with these changes.

Summary by CodeRabbit

• Bug Fixes
  • Improved graceful shutdown handling across startup paths to properly catch interrupts, log the interruption, and ensure clean termination of the application and web front-end.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Walkthrough

Adds KeyboardInterrupt handling to the NAT CLI start command and refactors the FastAPI front-end startup from blocking uvicorn.run() to creating a uvicorn.Config and uvicorn.Server and awaiting server.serve(), with explicit interrupt handling and graceful shutdown logging.

Changes

Cohort / File(s)	Summary
CLI interrupt handling src/nat/cli/commands/start.py	Added KeyboardInterrupt handling in run_plugin: logs the interruption and returns None.
FastAPI server interrupt handling src/nat/front_ends/fastapi/fastapi_front_end_plugin.py	Replaced uvicorn.run(...) with uvicorn.Config + uvicorn.Server and await server.serve(). Added try/except KeyboardInterrupt around serve() to log and shut down gracefully; preserved existing log formatting and Gunicorn branch.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly describes the main change: fixing event loop handling for Python 3.11 in the serve command, which aligns with the KeyboardInterrupt handling and uvicorn server refactoring in the changeset.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[willkill07]

/ok to test 588fe18

[yczhang-nv]

I like the fix - well-scoped and easy to understand

[coderabbitai]

Actionable comments posted: 1

🤖 Fix all issues with AI agents

In `@src/nat/front_ends/fastapi/fastapi_front_end_plugin.py`:
- Around line 202-218: The code currently constructs uvicorn.Config and calls
server.serve() which ignores multiprocess/reload; add a pre-check in the method
that builds the uvicorn.Config (the code block that reads self.front_end_config
and creates uvicorn.Config / server) to validate that if
self.front_end_config.workers > 1 or self.front_end_config.reload is True and
use_gunicorn (or the flag that controls using Gunicorn) is False, you reject the
combination: raise a clear ValueError (or log an error and exit) explaining that
workers>1 and reload=True require the Gunicorn/supervisor path and instruct the
user to enable use_gunicorn or set workers=1 and reload=False; reference
self.front_end_config.workers, self.front_end_config.reload, and the
uvicorn.Server(config).serve() call so the guard runs before creating
uvicorn.Server.

[coderabbitai]

Actionable comments posted: 1

🤖 Fix all issues with AI agents

In `@src/nat/front_ends/fastapi/fastapi_front_end_plugin.py`:
- Around line 214-217: In the except KeyboardInterrupt block that wraps await
server.serve(), replace the logger.info call with logger.exception (or
logger.error with exc_info=True) so the KeyboardInterrupt stack trace is
recorded; locate the try/except around server.serve() in
fastapi_front_end_plugin and change the logging call in the except
KeyboardInterrupt: handler to logger.exception("Received interrupt, shutting
down FastAPI server.") to preserve exception details.

[willkill07]

/merge