When connection queue is full, respond with a 503 error.

[itamarst]

❓ What kind of change does this PR introduce?

• 🐞 bug fix
• 🐣 feature
• 📋 docs update
• 📋 tests/coverage improvement
• 📋 refactoring
• 💥 other

❓ What is the current behavior? (You can also link to an open issue here)

When connections can't be handled, the connection is just closed.

❓ What is the new behavior (if this is a feature change)?

A 503 error is returned, as suggested in TODO comment in relevant part of the code.

Since writing to the socket is blocking, this is done in a thread.

📋 Other information:

For some reason OpenStack Ironic needs this apparently, I didn't dig into why.

📋 Contribution checklist:

(If you're a first-timer, check out
this guide on making great pull requests)

• I wrote descriptive pull request text above
• I think the code is well written
• I wrote good commit messages
• I have squashed related commits together after
  the changes have been approved
• Unit tests for the changes exist
• Integration tests for the changes exist (if applicable)
• I used the same coding conventions as the rest of the project
• The new code doesn't generate linter offenses
• Documentation reflects the changes
• The PR relates to only one subject with a clear title
  and description in grammatically correct, complete sentences

---

This change is 

[itamarst]

So now it ... fails, but not as consistently? I don't really know.

[itamarst]

Any other thoughts? It's pretty annoying that it passes some of the time but not all the time.

[itamarst]

Looks like we're down to only 3.14 failing, which were already failing on main so hopefully unrelated to my changes.

[itamarst]

Failures are due to 3.14 and #747, unrelated to this PR as far as I can tell.

[webknjaz]

@itamarst i've hit rebase on GH UI post a few CI fixes to main and it looks like the new test is flaky in there: https://github.com/cherrypy/cheroot/actions/runs/15811177788/job/44562688410?pr=745#step:15:1574

[itamarst]

Maybe that's why the linger was in the code I copied from. I'll test that out next week.

[itamarst]

Seems better maybe with linger added back?

[itamarst]

@webknjaz back to you!

[webknjaz]

@itamarst one thing I forgot to ask for is adding a change note for Towncrier. Could you do that?
https://cheroot.cherrypy.dev/en/latest/contributing/guidelines/#adding-change-notes-with-your-prs

[webknjaz]

@itamarst I think this is mostly ready, short of the requests above. When writing a change note, https://cheroot--745.org.readthedocs.build/en/745/history/#to-be-included-in-the-next-release can be used to preview rendering and compare the writing style with the surrounding text.

[itamarst]

OK back to you @webknjaz

[webknjaz]

Wonderful! Thanks for the contrib.

[webknjaz]

Looks like we have a regression coming out of this change: #769 / #778

[martin-vi]

There is no self.server within HTTPServer.